{ "schema_version": "1.14", "updated": "2026-10-07", "total": 494, "models": { "seedance-2.0": { "name": "Doubao Seedance 2.0", "vendor": "字节跳动 / 火山方舟", "max_duration": "15s", "max_input_videos": 3, "prompt_length": "50-200 词", "audio": "需单独处理", "strengths": [ "复杂叙事", "多主体", "电影感", "镜头切换" ] }, "happyhorse-1.0": { "name": "HappyHorse 1.0", "vendor": "阿里巴巴", "max_duration": "15s (千问 App 5/10/15s 三档,默认 5s)", "prompt_length": "30-55 词 (最优)", "audio": "原生联合生成音视频", "strengths": [ "⭐ 2026 年 5 月 Elo 排名第一(1357)", "紧凑短片", "ASMR", "原生带音频", "8s 时序节拍" ] }, "kling-3.0": { "name": "Kling 3.0 (可灵)", "vendor": "快手", "max_duration": "15s (单段) / 6 shots within 15s / 智能分镜 2m", "prompt_length": "中英文自由文本", "audio": "原生音画同步 + 角色定向发声", "strengths": [ "中文理解强", "物理感动作", "首尾帧", "图生视频", "Motion Brush", "智能分镜" ] }, "sora-2": { "name": "Sora 2", "vendor": "OpenAI", "max_duration": "25s (Pro) / 15s (普通)", "prompt_length": "80-150 词 (最优)", "audio": "原生音画 + 对话生成", "strengths": [ "电影质感", "Shot List 分层结构", "Cameos 真人客串", "极致物理仿真", "技术参数化提示词" ], "notice": "⚠️ OpenAI Videos API(含 sora-2 / sora-2-pro)已于 2026-09-24 停服,无替代模型;web/app 早前已于 2026-04-26 停止。Microsoft Foundry 最后一个部署 sora-2 (2025-12-08) 将于 2026-10-15 退役。新作品建议迁移到 Veo 3.1 / Seedance 2.5 / Wan 3.0" }, "veo-3": { "name": "Veo 3.1", "vendor": "Google DeepMind", "max_duration": "8s 单段 (Veo 3.1) / 148s (chained extend)", "prompt_length": "100-150 词 (3-6 句最优)", "audio": "原生音频 + 多人对话 + 精确音效同步", "strengths": [ "原生音频最强", "多人对话", "精确音效同步", "Storyboard 思维", "Google Cloud API" ] }, "runway-gen4": { "name": "Runway Gen-4.5 / Aleph", "vendor": "Runway", "max_duration": "10s (Gen-4.5) / 5s (Aleph 单次编辑)", "prompt_length": "自由文本,Aleph 用编辑动词", "audio": "需后期", "strengths": [ "Gen-4.5 ELO 排名第一", "视频编辑 Aleph 独家", "电影质感", "Motion Brush 引导", "API 接入方便" ] }, "pika-2.5": { "name": "Pika 2.5", "vendor": "Pika Labs", "max_duration": "25s(场景延长)", "prompt_length": "简短聚焦,单一主体", "audio": "部分支持", "strengths": [ "Pikaffects 15+ 创意特效", "iOS 原生 App", "短视频爆款", "单主体强一致性" ] }, "hailuo-02": { "name": "Hailuo 02 (海螺)", "vendor": "MiniMax", "max_duration": "10s", "prompt_length": "克制简洁,避免堆砌", "audio": "需后期", "strengths": [ "物理仿真业界第一", "1080p", "Director's AI 叙事思维", "重力/惯性/材质准确" ], "notice": "MiniMax 同期还有 Hailuo 2.3(性价比版本,约竞品一半价格)" }, "hunyuan-video": { "name": "Hunyuan Video 1.5", "vendor": "腾讯", "max_duration": "5-10s", "prompt_length": "Normal/Master 双模式 prompt rewrite", "audio": "需后期", "strengths": [ "开源最强(13B / 1.5 仅 8.3B)", "跨维度生成", "本地部署", "LoRA 训练友好" ] }, "wan-2.5": { "name": "Wan 2.7 (通义万相)", "vendor": "阿里巴巴", "max_duration": "15s (Wan 2.7)", "prompt_length": "Entity + Scene + Motion + Sound 公式", "audio": "原生音视频同步", "strengths": [ "原生音频", "数字人 lip-sync 业界最准", "对口型 / 广告 / MV", "Wan 2.6 开源 / 2.7 最新", "Alibaba Cloud API" ], "notice": "Wan 2.6 是稳定开源版(RTX 4090 可跑),Wan 2.7 是最新发布版(更长 prompt 窗口 + 更锐利控制)" }, "jimeng-3": { "name": "即梦 AI (Seedance 2.0 引擎)", "vendor": "字节跳动 / 剪映", "max_duration": "15s (基于 Seedance 2.0)", "prompt_length": "中文友好,8 维度公式", "audio": "原生音画同步 + 8+ 语言唇形对齐", "strengths": [ "中文最强(与 Kling 并列)", "剪映原生集成", "百款内置模板", "C 端创作者首选", "底层即 Seedance 2.0" ], "notice": "即梦是字节剪映 C 端产品,底层引擎即 Seedance 2.0(同源不同前端)" }, "ltx-video": { "name": "LTX-Video 0.9.7", "vendor": "Lightricks (开源)", "max_duration": "10s", "prompt_length": "4-8 句单段自然描述,200 词以内", "audio": "无原生", "strengths": [ "实时生成 (30 FPS 1216×704)", "DiT 架构最快", "单段落 prose 优先", "本地部署友好", "开源 (Apache 2.0)", "screenplay-style 对话场景" ] }, "mochi-1": { "name": "Mochi 1", "vendor": "Genmo (开源)", "max_duration": "5.4s", "prompt_length": "自由文本,强 prompt 遵循", "audio": "无原生", "strengths": [ "10B 参数 AsymmDiT", "已发布最大开源视频模型", "强 prompt 遵循", "30 FPS 流畅运动", "Apache 2.0 开源", "Genmo Playground 浏览器可试" ] }, "cogvideox-5b": { "name": "CogVideoX 5B / 1.5", "vendor": "智谱 AI · 清华 (开源)", "max_duration": "10s", "prompt_length": "英文最多 226 token,详细场景", "audio": "无原生", "strengths": [ "768×1360 高分辨率", "5B 参数轻量", "中国出品开源", "支持 T2V + I2V", "HuggingFace 完整工作流", "Apache 2.0 开源" ] }, "higgsfield-soul": { "name": "Higgsfield Soul 2.0 / DoP", "vendor": "Higgsfield AI", "max_duration": "15s (单段) / 60-90s (AI Director 智能分镜)", "prompt_length": "短指令 > 长描述 + 模板系统", "audio": "原生支持", "strengths": [ "Soul ID 角色一致性最强", "DoP 电影镜头模板", "社交媒体 viral 短片", "AI Director 智能分镜", "20+ Sub-skills" ] }, "gemini-omni": { "name": "Gemini Omni (Omni Flash)", "vendor": "Google AI / DeepMind", "max_duration": "(待官方确认 · 已上线 Gemini App / Flow / YouTube Shorts/Create)", "prompt_length": "灵活,可用文化/科学锚点替代细节描述", "audio": "原生多模态(any-input-to-anything)", "strengths": [ "跨模态创作 (any input → any output)", "实时文字渲染 (typography / 空间放置 / 动画风格)", "迭代编辑 (targeted updates 不重写)", "中途动作修改 (修改 pacing/emotion 保持连续性)", "Real-World 知识利用 (文化/历史/科学锚点直接用)", "Gemini App / Flow / YouTube Shorts 多平台" ], "notice": "2026-05-19 发布 · 跨模态统一模型,从视频开始(Google AI 官方介绍 5 个 prompting tips)" }, "kling-4.0": { "name": "Kling 4.0 (可灵)", "vendor": "快手", "max_duration": "30s(官方规格,原生单次;实验性 120s 长视频模式)", "prompt_length": "结构化长文本([OUTPUT]/[REFERENCE ROLES]/[CONTINUITY ANCHORS] 区块)", "audio": "原生对白 + 9 语言 lip sync", "strengths": [ "omni reference 多引用叙事", "match-cut 藏转场", "多语言对白口型同步" ] }, "minimax-h3": { "name": "MiniMax H3", "vendor": "MiniMax", "max_duration": "15s(官方规格 4–15s)", "prompt_length": "短文本 + 音视频统一 reference", "audio": "原生音频:角色音色授权 + 环境音分离 + 口型自动同步", "strengths": [ "音视频统一 reference", "精准台词生成", "多角色音色区分" ] }, "seedance-2.5": { "name": "Seedance 2.5(即梦)", "vendor": "字节跳动", "max_duration": "30s(官方规格 4–30s,可延长)", "prompt_length": "中英文自由文本", "audio": "原生对白 + 配乐时间轴", "strengths": [ "微表情连续变化", "一镜到底叙事", "情绪表演" ] }, "wan-2.6": { "name": "Wan 2.6(万相)", "vendor": "阿里", "max_duration": "见官方 R2V 文档", "prompt_length": "多 reference 联合驱动描述", "audio": "多人对白", "strengths": [ "reference-to-video(R2V)", "视频+图片+音频多引用", "多人对白表演" ] }, "runway-aleph-2.0": { "name": "Runway Aleph 2.0", "vendor": "Runway", "max_duration": "见官方文档", "prompt_length": "变化描述 + anchor frame reference", "audio": "—", "strengths": [ "anchor-frame 精确编辑", "只命名变化、首帧定外观", "品牌级精确替换" ] }, "luma-ray-3.2": { "name": "Luma Ray 3.2", "vendor": "Luma AI", "max_duration": "见官方文档", "prompt_length": "shot-direction 公式(机位+主体+动作+场景+光影+风格+一致性约束)", "audio": "—", "strengths": [ "shot-direction 公式", "单动作聚焦", "一致性约束写法" ] }, "wan-3.0": { "name": "Wan 3.0(通义万相)", "vendor": "阿里通义实验室", "max_duration": "30s(官方规格,单次生成)", "prompt_length": "文本/图片/音频/视频/文档(PDF/PPT/表格/网页)多模态输入", "audio": "原生音频:多语言配音 + 表情同步", "strengths": [ "30s 单次生成", "文档转视频(PDF/PPT/表格/网页)", "多模态 reference 输入" ], "notice": "2026-08-24 正式发布(08-06 公测);API 按秒计费" } }, "categories": [ { "id": "product-commercial", "zh": "产品与商业广告", "model": "both" }, { "id": "social-viral", "zh": "社交媒体爆款", "model": "seedance-2.0" }, { "id": "cinematic", "zh": "电影叙事与戏剧场景", "model": "seedance-2.0" }, { "id": "action", "zh": "动作、格斗与追逐", "model": "both" }, { "id": "sports", "zh": "体育、赛车与武术", "model": "seedance-2.0" }, { "id": "nature", "zh": "自然、风景与天气", "model": "both" }, { "id": "quiet-moments", "zh": "安静时刻与生活切片", "model": "seedance-2.0" }, { "id": "fantasy-scifi", "zh": "奇幻与科幻", "model": "seedance-2.0" }, { "id": "creative", "zh": "创意与实验性", "model": "seedance-2.0" }, { "id": "asmr", "zh": "ASMR 与感官体验", "model": "seedance-2.0" }, { "id": "comedy", "zh": "喜剧与概念性", "model": "seedance-2.0" }, { "id": "craft", "zh": "工艺、艺术与现场表演", "model": "seedance-2.0" }, { "id": "cinematic-timed", "zh": "时序节拍电影感", "model": "happyhorse-1.0" }, { "id": "portrait", "zh": "人物与肖像", "model": "happyhorse-1.0" }, { "id": "cinematic-styles", "zh": "电影风格", "model": "happyhorse-1.0" }, { "id": "viral-transform", "zh": "病毒变身与转场", "model": "kling-3.0" }, { "id": "music-video", "zh": "音乐 MV 与表演", "model": "kling-3.0" }, { "id": "experimental", "zh": "实验与超现实", "model": "kling-3.0" }, { "id": "gaming-ip", "zh": "游戏与 IP 风", "model": "kling-3.0" }, { "id": "image-to-video", "zh": "图生视频专项", "model": "kling-3.0" }, { "id": "horror", "zh": "恐怖与悬疑", "model": "both" }, { "id": "documentary", "zh": "纪录片与采访", "model": "both" }, { "id": "dialogue-driven", "zh": "对话驱动(原生音频)", "model": "both" }, { "id": "video-edit", "zh": "视频编辑与重生(Aleph)", "model": "runway-gen4" }, { "id": "pikaffects", "zh": "Pikaffects 创意特效", "model": "pika-2.5" }, { "id": "physics-sim", "zh": "物理仿真精确动作", "model": "hailuo-02" }, { "id": "open-source", "zh": "开源本地部署专项", "model": "hunyuan-video" }, { "id": "lip-sync", "zh": "数字人对口型", "model": "wan-2.5" }, { "id": "chinese-native", "zh": "中文原生创作", "model": "both" }, { "id": "technique-snippet", "zh": "技巧片段与提示词语法", "model": "both" } ], "prompts": [ { "id": "cg-001", "model": "cogvideox-5b", "category": "portrait", "title": "CogVideoX 东京时尚漫步", "prompt": "A stylish woman walks down a Tokyo street filled with warm glowing neon and animated city signage. She wears a black leather jacket, a long red dress, and black boots, and carries a black purse. She wears sunglasses and red lipstick. She walks confidently and casually. The street is damp and reflective, creating a mirror effect of the colorful lights, and many pedestrians walk about.", "tags": [ "portrait", "tokyo", "neon", "cogvideo-official" ], "params": { "aspect": "16:9", "duration": "10s", "resolution": "1360x768" }, "source": { "name": "Zhipu CogVideoX", "url": "https://github.com/zai-org/CogVideo", "tier": "official-platform" }, "notes": "CogVideoX 官方样板 prompt,英文最多 226 token,详细场景描述" }, { "id": "cg-002", "model": "cogvideox-5b", "category": "nature", "title": "CogVideoX 海岸日落浪潮", "prompt": "Ocean waves crashing against a rocky coastline at sunset. Spray and mist catching the golden light. Camera positioned low, watching waves approach and recede. Natural sound implied, peaceful yet powerful atmosphere. The horizon glows with deep orange and pink hues, seabirds occasionally drift across the frame, and the rhythm of the waves dominates the composition.", "tags": [ "nature", "ocean", "sunset", "cogvideo-official" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Zhipu", "url": "https://huggingface.co/zai-org/CogVideoX-5b", "tier": "official-platform" } }, { "id": "cg-003", "model": "cogvideox-5b", "category": "comedy", "title": "CogVideoX 弹吉他熊猫", "prompt": "A panda, dressed in a small, red jacket and a tiny hat, sits on a wooden stool in a serene bamboo forest. The panda's fluffy paws strum a miniature acoustic guitar, producing soft, melodic tunes. Nearby, a few other pandas gather, watching curiously and some clapping in rhythm. Sunlight filters through the tall bamboo, casting a gentle glow on the scene. The panda's face is expressive, showing concentration and joy as it plays.", "tags": [ "comedy", "animal", "music", "cogvideo-official" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Zhipu", "url": "https://github.com/zai-org/CogVideo", "tier": "official-platform" } }, { "id": "cg-004", "model": "cogvideox-5b", "title": "CogVideoX I2V 风景动态化", "prompt": "Based on the provided still image of a tranquil landscape, gentle motion brings the scene to life over the course of six seconds. Water in the foreground ripples softly as a light breeze passes across its surface, concentric circles spreading outward from a single point. Leaves on the trees lining the shore sway with the wind, their fresh green coloring catching shifts of dappled sunlight. Distant clouds drift slowly across a brilliantly clear sky, their shadows moving subtly across the rolling hills behind. The camera holds a steady wide composition, capturing the serene movement of nature without any cuts. Soft, diffused daylight bathes the entire scene in warm golden tones, creating a picturesque, contemplative atmosphere.", "tags": [ "image-to-video", "nature", "ambient", "cogvideo" ], "category": "image-to-video", "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "Zhipu CogVideoX-5b-I2V", "url": "https://huggingface.co/zai-org/CogVideoX-5b-I2V", "tier": "official-platform" } }, { "id": "cg-005", "model": "cogvideox-5b", "title": "CogVideoX 本地部署测试 prompt", "prompt": "A cat with bright green eyes sits on a wooden windowsill, watching raindrops trace winding patterns down the cold glass. The cat's striped grey fur is illuminated by the warm interior light, while outside, blurred lights of a city evening glow softly in the rain-soaked street. The cat's tail twitches occasionally, ears rotating slightly toward the sound of distant traffic. Each raindrop catches a tiny reflection of the city lights as it slides down the pane. The camera holds a static medium close-up framing the cat in profile against the window, capturing the contrast between the cozy indoor warmth and the cool blue tones of the rainy night outside.", "tags": [ "pet", "rainy", "test", "open-source" ], "category": "open-source", "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Zhipu CogVideoX", "url": "https://github.com/zai-org/CogVideo", "tier": "official-platform" } }, { "id": "cg-006", "model": "cogvideox-5b", "category": "cinematic", "title": "CogVideoX 上海雨夜小巷", "prompt": "A narrow alley in old Shanghai at night, drizzle falling steadily, neon signs in Chinese characters reflected on wet cobblestones. An elderly tea vendor stands beside his small wooden cart under a single hanging bulb, pouring tea into a small ceramic cup. Steam rises from the cup. The camera slowly pushes in from behind a parked bicycle, framing the vendor and his cart in soft warm light against the cool blue night.", "tags": [ "shanghai", "noir", "rain", "atmospheric" ], "params": { "aspect": "21:9", "duration": "10s" }, "source": { "name": "CogVideoX Community", "url": "https://github.com/zai-org/CogVideo", "tier": "official-platform" } }, { "id": "cg-007", "model": "cogvideox-5b", "title": "CogVideoX 手表精致展示", "prompt": "A close-up of a luxury mechanical watch rests on a polished walnut surface, the rich grain of the wood catching warm directional light. The second hand sweeps smoothly across the white enamel dial, each tick rendered in fine detail. The camera slowly orbits around the watch, revealing the brushed-steel case from multiple angles. Through the transparent caseback in alternating shots, the intricate movement of internal gears, jewels, and the balance wheel becomes visible, each component working in delicate synchronization. Single key light from the upper right creates crisp specular highlights along the polished bezel and crown, while soft shadows fall behind the watch on the wood. The atmosphere is serene and contemplative, highlighting the artistry of the timepiece.", "tags": [ "product", "watch", "macro", "luxury" ], "category": "product-commercial", "params": { "aspect": "1:1", "duration": "8s" }, "source": { "name": "Zhipu CogVideoX", "url": "https://github.com/zai-org/CogVideo", "tier": "official-platform" } }, { "id": "cg-008", "model": "cogvideox-5b", "category": "experimental", "title": "CogVideoX 流光抽象艺术", "prompt": "An abstract composition of flowing colored liquids merging and separating against a pure black background, deep indigo and electric coral colors swirling in intricate fractal patterns, slow rotating motion creating mesmerizing visual rhythm, hyperreal liquid surface tension with metallic sheen catching directional light, satisfying loop.", "tags": [ "abstract", "fluid", "art", "loop" ], "params": { "aspect": "1:1", "duration": "10s" }, "source": { "name": "CogVideoX", "url": "https://huggingface.co/zai-org/CogVideoX-5b", "tier": "official-platform" } }, { "id": "gm-001", "model": "gemini-omni", "title": "26 字母物品节拍剪辑(Google AI 官方样板)", "prompt": "The video shows items of the alphabet. An unusual item starting with each letter is shown sitting on a table (like a Capybara for C, disco globe for D and Lava Lamp for L). All 26 letters must be represented by 26 items with matching lower thirds displaying the letter. Only one item and lower third at a time. Each lower third must look like a black marker written on a slip of paper in the bottom left. Rapid fire, roughly 9 frames per item at 24FPS. Last frame is a slip of paper \"THE END.\" The whole video is accompanied by calm smooth music.", "tags": [ "text-rendering", "rapid-fire", "alphabet", "educational", "google-official" ], "category": "creative", "params": { "aspect": "16:9", "duration": "~10s", "resolution": "1080p" }, "source": { "name": "Google AI · Mastering Gemini Omni: The Ultimate Video Prompting Guide", "url": "https://x.com/GoogleAI/status/2059381218660270435", "tier": "official" } }, { "id": "gm-002", "model": "gemini-omni", "title": "极简 prompt:火星宇航员视角", "prompt": "Astronaut's POV on Mars", "tags": [ "minimal-prompt", "pov", "space", "real-world-knowledge", "google-official" ], "category": "fantasy-scifi", "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google AI · Tip 1 Real-World Knowledge", "url": "https://x.com/GoogleAI/status/2059381218660270435", "tier": "official" } }, { "id": "gm-003", "model": "gemini-omni", "title": "Rube Goldberg 大理石链式反应(极简)", "prompt": "A marble rolling fast on a chain reaction style track, continuous smooth shot", "tags": [ "minimal-prompt", "physics", "marble", "continuous-shot", "google-official" ], "category": "creative", "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google AI · Tip 1 Real-World Knowledge", "url": "https://x.com/GoogleAI/status/2059381218660270435", "tier": "official" } }, { "id": "gm-004", "model": "gemini-omni", "title": "文字单词逐个显示(节拍 sizzle reel)", "prompt": "word by word, one word on the screen at a time: did, you, know, that, this, model, can, do, pretty, good, text!? Each word appears with a different animated style, perfect pacing to a rhythm, sizzle reel.", "tags": [ "text-rendering", "typography", "rhythm", "sizzle-reel", "google-official" ], "category": "creative", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Google AI · Tip 2 Text Rendering", "url": "https://x.com/GoogleAI/status/2059381218660270435", "tier": "official" } }, { "id": "gm-005", "model": "gemini-omni", "title": "内心独白漂浮 3D 文字(Intrusive Thoughts)", "prompt": "Overlay motion-tracked, minimalist text commentary onto the physical environment of the video. This text represents [the subject] deadpan, immediate inner monologue that's observant, slightly absurd, and life-contemplating. Think \"intrusive thoughts.\" Clean, white, lowercase sans-serif text (like Helvetica or Inter). The text hovers in 3D space, connected to the subjects being commented on via ultra-thin, crisp, white leader lines.", "tags": [ "text-rendering", "motion-tracked", "3d-text", "monologue", "absurd", "google-official" ], "category": "creative", "params": { "aspect": "16:9", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Google AI · Tip 2 Text Rendering (Intrusive Thoughts)", "url": "https://x.com/GoogleAI/status/2059381218660270435", "tier": "official" } }, { "id": "gm-006", "model": "gemini-omni", "title": "迭代编辑:把小提琴移到新环境", "prompt": "Transport the violin to a new environment", "tags": [ "iterative-edit", "environment-change", "minimal-prompt", "google-official" ], "category": "creative", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Google AI · Tip 4 Edit Iteratively", "url": "https://x.com/GoogleAI/status/2059381218660270435", "tier": "official" } }, { "id": "gm-007", "model": "gemini-omni", "title": "迭代编辑:小提琴隐形", "prompt": "Make the violin invisible", "tags": [ "iterative-edit", "vfx", "invisibility", "minimal-prompt", "google-official" ], "category": "creative", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Google AI · Tip 4 Edit Iteratively", "url": "https://x.com/GoogleAI/status/2059381218660270435", "tier": "official" } }, { "id": "gm-008", "model": "gemini-omni", "title": "迭代编辑:换镜头角度(over-shoulder)", "prompt": "Change the camera angle so it's looking over the violinist's shoulder", "tags": [ "iterative-edit", "camera-angle", "over-shoulder", "minimal-prompt", "google-official" ], "category": "creative", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Google AI · Tip 4 Edit Iteratively", "url": "https://x.com/GoogleAI/status/2059381218660270435", "tier": "official" } }, { "id": "gm-009", "model": "gemini-omni", "title": "动作即时修改:踮脚走", "prompt": "Make the character walk on their tiptoes", "tags": [ "action-modify", "character-control", "minimal-prompt", "google-official" ], "category": "creative", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Google AI · Tip 5 Change Action on the Fly", "url": "https://x.com/GoogleAI/status/2059381218660270435", "tier": "official" } }, { "id": "gm-010", "model": "gemini-omni", "title": "动作即时修改:跳起", "prompt": "Have them leap into the air", "tags": [ "action-modify", "character-control", "minimal-prompt", "google-official" ], "category": "creative", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Google AI · Tip 5 Change Action on the Fly", "url": "https://x.com/GoogleAI/status/2059381218660270435", "tier": "official" } }, { "id": "hl-001", "model": "hailuo-02", "category": "physics-sim", "title": "海浪冲击礁石(物理仿真)", "prompt": "Massive ocean waves crash against rugged cliff rocks at high speed, water spray erupts upward and disperses in mist, foam pools form and drain in cycles. Wide shot from sea level, late afternoon backlight catching the spray. Realistic fluid dynamics, accurate gravity and water tension.", "tags": [ "physics", "ocean", "fluid", "atmospheric" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "MiniMax", "url": "https://platform.minimax.io/docs/guides/video-generation", "tier": "official" }, "notes": "Hailuo 02 物理仿真业界第一,水/火/烟/布料尤其准确" }, { "id": "hl-002", "model": "hailuo-02", "title": "烛火被风吹动(火焰物理)", "prompt": "A single candle flame flickers and bends as a gust of wind passes through a dark room. The flame stretches sideways with accurate fluid dynamics, smoke trails rise in twisting columns, hot wax pools and drips slowly down the side of the candle. Extreme close-up, side-light only from the flame itself.", "tags": [ "physics", "fire", "candle", "macro" ], "category": "physics-sim", "params": { "aspect": "1:1", "duration": "6s", "resolution": "1080p" }, "source": { "name": "MiniMax Platform", "url": "https://platform.minimax.io/docs/guides/video-generation", "tier": "official" } }, { "id": "hl-003", "model": "hailuo-02", "category": "physics-sim", "title": "丝绸在风中飘动(布料物理)", "prompt": "A long red silk scarf floats and ripples in the wind against a clear blue sky, fabric folds and stretches with realistic weight, sunlight catches every wave. Medium shot from below, the scarf fills most of the frame. Slow graceful motion.", "tags": [ "physics", "fabric", "silk", "slow-motion" ], "params": { "aspect": "9:16", "duration": "8s", "resolution": "1080p" }, "source": { "name": "MiniMax", "url": "https://platform.minimax.io/docs/guides/video-generation", "tier": "official" } }, { "id": "hl-004", "model": "hailuo-02", "category": "physics-sim", "title": "体操运动员空中翻转(重力惯性)", "prompt": "A female gymnast performs a complex aerial flip on a balance beam, body rotates with accurate angular momentum, hair and leotard catch wind, lands cleanly with knee absorption. Side-angle wide shot, gym lighting, slow motion at apex.", "tags": [ "physics", "sports", "gymnastics", "slow-motion" ], "params": { "aspect": "16:9", "duration": "10s", "resolution": "1080p" }, "source": { "name": "getimg", "url": "https://getimg.ai/blog/minimax-hailuo-02-the-1080p-ai-video-model-that-gets-physics-right", "tier": "reputable-3rd-party" } }, { "id": "hl-005", "model": "hailuo-02", "category": "physics-sim", "title": "陶瓷碗坠落破碎(脆性物理)", "prompt": "A blue ceramic bowl slips from a wooden table and falls, hits the floor and shatters into precise fragments, shards bounce and scatter outward, dust puff rises. Side-angle close-up, kitchen ambient light, real-time speed with slight slow-mo at impact.", "tags": [ "physics", "destruction", "ceramics", "impact" ], "params": { "aspect": "1:1", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Hailuo", "url": "https://blog.segmind.com/hailuo-minimax-ai-video-prompt-guide/", "tier": "reputable-3rd-party" } }, { "id": "hl-006", "model": "hailuo-02", "title": "篮球扣篮特写", "prompt": "A basketball player rises and slams a dunk through the hoop. The ball drops cleanly through the net, the rim and backboard vibrate from impact with accurate physics, the player hangs briefly on the rim before releasing. Low-angle hero shot from courtside, real-time playback, hardwood floor reflections.", "tags": [ "sports", "basketball", "low-angle", "slow-motion" ], "category": "action", "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "MiniMax Platform", "url": "https://platform.minimax.io/docs/guides/video-generation", "tier": "official" } }, { "id": "hl-007", "model": "hailuo-02", "title": "雪夜街头的拥抱", "prompt": "Two people in heavy coats embrace on a snowy city street at night under a streetlamp. Snowflakes settle on their shoulders and hair with accurate weight, breath fogs visibly in the cold air, fabric of their coats compresses where pressed together. Medium shot, slow circling camera, warm tungsten light against cold blue night.", "tags": [ "emotional", "winter", "street", "circling" ], "category": "cinematic", "params": { "aspect": "16:9", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Hailuo Prompt Guide", "url": "https://blog.segmind.com/hailuo-minimax-ai-video-prompt-guide/", "tier": "reputable-3rd-party" } }, { "id": "hl-008", "model": "hailuo-02", "category": "nature", "title": "蜂鸟悬停吸蜜(精细动作)", "prompt": "A hummingbird hovers in front of a red trumpet flower, wings blur at extreme speed, long beak extends into the flower. Macro close-up, soft afternoon sunlight, blurred green garden background. Real-time wing speed, ultra-realistic feather detail.", "tags": [ "wildlife", "macro", "hummingbird", "physics" ], "params": { "aspect": "16:9", "duration": "6s", "resolution": "1080p" }, "source": { "name": "Hailuo", "url": "https://platform.minimax.io/docs/guides/video-generation", "tier": "official" } }, { "id": "hl-009", "model": "hailuo-02", "title": "小狗失败接飞盘", "prompt": "A golden retriever puppy enthusiastically leaps to catch a frisbee but misses, tumbles into soft grass with accurate gravity, immediately bounces back up tail wagging. Fur ripples on impact, the disc spins past in the background. Side-angle medium shot, sunny park afternoon, real-time playback.", "tags": [ "pet", "comedy", "physical", "outdoor" ], "category": "comedy", "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "MiniMax Platform", "url": "https://platform.minimax.io/docs/guides/video-generation", "tier": "official" } }, { "id": "hl-010", "model": "hailuo-02", "category": "asmr", "title": "煎牛排油花飞溅(食物物理)", "prompt": "A thick beef steak hits a sizzling cast-iron pan, oil sprays violently outward, surface caramelizes and crusts. Extreme close-up, top-down angle, kitchen warm lighting. Audible aggressive sizzling sound. Realistic Maillard reaction.", "tags": [ "food", "asmr", "macro", "sizzle" ], "params": { "aspect": "1:1", "duration": "6s", "resolution": "1080p" }, "source": { "name": "Hailuo", "url": "https://blog.segmind.com/hailuo-minimax-ai-video-prompt-guide/", "tier": "reputable-3rd-party" } }, { "id": "hl-011", "model": "hailuo-02", "title": "沙漠弃车爆胎", "prompt": "An abandoned vintage car sits on a desolate desert road, its rear tire suddenly bursts. Rubber shreds fly outward in accurate angular momentum, the chassis settles with realistic suspension compression, dust kicks up and disperses in the dry wind. Wide low-angle shot, stormy dusk lighting, real-time motion.", "tags": [ "physics", "destruction", "desert", "cinematic" ], "category": "cinematic", "params": { "aspect": "21:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Hailuo Prompt Guide", "url": "https://blog.segmind.com/hailuo-minimax-ai-video-prompt-guide/", "tier": "reputable-3rd-party" } }, { "id": "hl-012", "model": "hailuo-02", "category": "experimental", "title": "墨水滴入清水(流体艺术)", "prompt": "A single drop of black ink falls into a glass of clear water, ink spreads in slow tendrils, forming intricate fractal patterns before fully diffusing. Extreme macro, side-backlight, pure black background. Hyperreal fluid dynamics.", "tags": [ "physics", "fluid", "macro", "art" ], "params": { "aspect": "1:1", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Hailuo", "url": "https://getimg.ai/blog/minimax-hailuo-02-the-1080p-ai-video-model-that-gets-physics-right", "tier": "reputable-3rd-party" } }, { "id": "hh-001", "model": "happyhorse-1.0", "category": "cinematic-timed", "title": "雨中面部特写(带时序)", "prompt": "8s duration. Extreme close-up of a woman in her 30s, dark wet hair, rain running down her face, eyes focused ahead, subtle jaw tension. Camera holds still. Shallow depth of field. Overcast grey light from directly in front. Slow-motion feel, realistic skin texture, cinematic realism.", "tags": [ "close-up", "rain", "static", "timed", "cinematic" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "CrePal", "url": "https://crepal.ai/blog/aivideo/aivideo-happyhorse-1-0-prompts/", "tier": "reputable-3rd-party" } }, { "id": "hh-002", "model": "happyhorse-1.0", "category": "cinematic-timed", "title": "广角建立镜头(带时序)", "prompt": "8s duration, first 2s: black. Slow fade reveals: wide shot, mountain valley at dawn, low mist between pine trees, single dirt road leading into the scene, no people. Camera very slowly pushes forward on a dolly. Soft blue-gold light on the horizon. Quiet, cinematic, high production value.", "tags": [ "wide", "establishing", "dawn", "timed", "fade-in" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "CrePal", "url": "https://crepal.ai/blog/aivideo/aivideo-happyhorse-1-0-prompts/", "tier": "reputable-3rd-party" } }, { "id": "hh-003", "model": "happyhorse-1.0", "category": "cinematic-timed", "title": "动作序列(带时序)", "prompt": "6s duration. A skateboarder lands a kick-flip on wet asphalt in an empty parking lot at night. Yellow sodium lights above. Low-angle side tracking shot moving with the board. Slow motion on impact, then returns to normal speed. Realistic motion blur. Urban, gritty, authentic.", "tags": [ "action", "skateboard", "night", "low-angle", "timed" ], "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "CrePal", "url": "https://crepal.ai/blog/aivideo/aivideo-happyhorse-1-0-prompts/", "tier": "reputable-3rd-party" } }, { "id": "hh-004", "model": "happyhorse-1.0", "category": "cinematic-timed", "title": "角色叙事(情绪驱动)", "prompt": "8s duration. A young man, early 20s, dark curly hair, wearing a faded denim jacket, sits alone at a diner table at night. Empty coffee cup in front of him. He looks out the window, then slowly back at the table. Close-medium shot. Warm tungsten interior light against cold dark exterior. Quiet, slightly melancholic, naturalistic.", "tags": [ "character", "narrative", "melancholic", "diner" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "CrePal", "url": "https://crepal.ai/blog/aivideo/aivideo-happyhorse-1-0-prompts/", "tier": "reputable-3rd-party" } }, { "id": "hh-005", "model": "happyhorse-1.0", "category": "cinematic-timed", "title": "情绪场景(行为表达)", "prompt": "7s duration. A woman in her 40s stands in a doorway, one hand on the frame. She looks down the hallway, pauses, then slowly closes the door. Medium shot, static camera. Soft warm interior light. No dialogue. The scene communicates loss without showing it directly.", "tags": [ "emotion", "subtle", "static", "doorway" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "CrePal", "url": "https://crepal.ai/blog/aivideo/aivideo-happyhorse-1-0-prompts/", "tier": "reputable-3rd-party" } }, { "id": "hh-006", "model": "happyhorse-1.0", "category": "product-commercial", "title": "产品英雄镜头(轨道运动)", "prompt": "8s duration. A luxury skincare serum bottle on a black marble surface, soft studio key light from upper left, subtle specular highlight along the glass edge. Camera slowly orbits the bottle in a 60-degree arc. Clean minimal aesthetic. Product stays sharp throughout. Commercial quality.", "tags": [ "product", "skincare", "orbit", "studio" ], "params": { "aspect": "1:1", "duration": "8s" }, "source": { "name": "CrePal", "url": "https://crepal.ai/blog/aivideo/aivideo-happyhorse-1-0-prompts/", "tier": "reputable-3rd-party" } }, { "id": "hh-007", "model": "happyhorse-1.0", "category": "cinematic-timed", "title": "生活方式场景(活力女性)", "prompt": "7s duration. A woman in activewear stands on a rooftop at sunrise, facing the city, holding a water bottle. She takes a sip, looks out. Light tracking shot moving from side to behind. Golden hour warm light. Aspirational, clean, athletic lifestyle feel. No text. No logo.", "tags": [ "lifestyle", "fitness", "rooftop", "golden-hour" ], "params": { "aspect": "9:16", "duration": "7s" }, "source": { "name": "CrePal", "url": "https://crepal.ai/blog/aivideo/aivideo-happyhorse-1-0-prompts/", "tier": "reputable-3rd-party" } }, { "id": "hh-008", "model": "happyhorse-1.0", "category": "product-commercial", "title": "智能手表英语产品讲解(原生音频)", "prompt": "A male product presenter in a clean white studio holds up a slim black smartwatch, turning it slowly to show the side profile. He explains the display features, bright even key lighting, minimal background, 16:9 wide shot, speaking English at a natural pace.", "tags": [ "product", "presenter", "smartwatch", "studio", "english", "native-audio" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happyhorse-1-0-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-009", "model": "happyhorse-1.0", "category": "social-viral", "title": "巴黎街拍秋日时装大片(带音频)", "prompt": "A French fashion model walks a sunlit cobblestone street in Paris, wearing a structured camel coat. Slow tracking shot from behind, transitioning to a frontal medium shot as she turns. Golden hour lighting, with heels on stone and distant city ambiance audible.", "tags": [ "fashion", "paris", "tracking", "golden-hour", "footsteps" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happyhorse-1-0-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-010", "model": "happyhorse-1.0", "category": "cinematic-styles", "title": "日语课堂太阳系讲解(原生音频)", "prompt": "A young male teacher stands at a whiteboard, drawing a simple diagram of the solar system with a marker. He explains each planet in Japanese, gesturing toward the drawing. Bright classroom lighting, medium wide shot, speaking Japanese at a measured pace with marker squeaks audible.", "tags": [ "education", "japanese", "classroom", "teacher", "native-audio" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happyhorse-1-0-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-011", "model": "happyhorse-1.0", "category": "asmr", "title": "抹茶冲泡 ASMR 特写(带音频)", "prompt": "Close-up of a woman's hands preparing matcha in a traditional ceramic bowl, whisking the powder into a smooth green paste. No dialogue, natural daylight from a side window, extremely shallow depth of field, with the soft sound of the whisk audible in ambient silence.", "tags": [ "asmr", "matcha", "close-up", "hands", "silence" ], "params": { "aspect": "1:1", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happyhorse-1-0-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-012", "model": "happyhorse-1.0", "category": "dialogue-driven", "title": "普通话企业发言人产品发布(原生音频)", "prompt": "A male spokesperson in his 40s sits in a modern office, speaking confidently to camera about a new product launch. Sharp corporate lighting, shallow background blur, tight mid-shot framing, professional authoritative tone, speaking Mandarin at a natural pace.", "tags": [ "dialogue", "mandarin", "corporate", "spokesperson", "native-audio" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happyhorse-1-0-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-013", "model": "happyhorse-1.0", "category": "action", "title": "晨雾公园跑步(带呼吸音)", "prompt": "A woman in athletic wear runs through a misty early morning park, her breath visible in the cold air. The camera tracks alongside her at eye level. Desaturated cool tones, cinematic wide aspect ratio, with footsteps, breathing, and distant birdsong audible.", "tags": [ "action", "running", "park", "mist", "tracking", "breathing" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happyhorse-1-0-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-014", "model": "happyhorse-1.0", "category": "craft", "title": "精品咖啡馆浓缩萃取(带音频)", "prompt": "A barista in a specialty coffee shop pulls an espresso shot, the crema swirling in slow motion as it fills a small glass. Low warm lighting, extreme close-up transitioning to a medium shot of the barista looking up with a slight smile, with cafe ambient sounds audible.", "tags": [ "food-beverage", "coffee", "espresso", "barista", "slow-motion" ], "params": { "aspect": "1:1", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happyhorse-1-0-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-015", "model": "happyhorse-1.0", "category": "dialogue-driven", "title": "德语录音棚配音演员(原生音频)", "prompt": "A German voiceover artist records in a professional studio booth, speaking calmly into a large condenser microphone. Tight mid-shot through the glass partition, studio monitor glow in the background, speaking German at a natural pace with clean studio audio.", "tags": [ "dialogue", "german", "studio", "voiceover", "native-audio" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happyhorse-1-0-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-016", "model": "happyhorse-1.0", "category": "cinematic-styles", "title": "粤语海崖黄昏低语(带环境音)", "prompt": "A couple walks along a windswept coastal cliff at dusk, speaking quietly to each other in Cantonese. Wide establishing shot pulling into a medium two-shot. Fading golden light, long shadows, speaking Cantonese at a measured pace with wind and breaking waves audible.", "tags": [ "narrative", "cantonese", "coast", "dusk", "couple", "native-audio" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happyhorse-1-0-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-017", "model": "happyhorse-1.0", "category": "portrait", "title": "渔夫码头补网晨光", "prompt": "An elderly fisherman mends a net on a weathered dock at dawn, slow dolly in from wide to close-up on his hands, warm low morning light, soft mist over the water, with gentle harbour water and rope tension audible.", "tags": [ "portrait", "fisherman", "dawn", "dolly", "hands" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-018", "model": "happyhorse-1.0", "category": "portrait", "title": "咖啡馆雨窗少女静观", "prompt": "A teenage girl sits at a cafe window watching rain, steam rising from a mug in front of her, locked-off static mid-shot, overcast grey light, shallow depth of field, with light rain on the window audible.", "tags": [ "portrait", "cafe", "rain", "static", "teen" ], "params": { "aspect": "9:16", "duration": "6s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-019", "model": "happyhorse-1.0", "category": "portrait", "title": "空荡舞厅暗色西装男士", "prompt": "A man in a dark suit stands alone in an empty ballroom, looks slowly toward camera, slight smile, steadicam push from wide to mid-shot, warm tungsten chandeliers, late evening, with faint hall reverb audible.", "tags": [ "portrait", "suit", "ballroom", "steadicam", "tungsten" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-020", "model": "happyhorse-1.0", "category": "portrait", "title": "读信特写表情转变", "prompt": "Close-up of a woman's face as she reads a letter, expression shifting from neutral to stunned, locked-off static, single soft window light from the left, shallow depth of field, with quiet paper rustle audible.", "tags": [ "portrait", "close-up", "letter", "emotion", "window-light" ], "params": { "aspect": "9:16", "duration": "6s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-021", "model": "happyhorse-1.0", "category": "action", "title": "厨房主厨摆盘跟手特写", "prompt": "A young chef plates a dish with focused precision in a busy kitchen, mid tracking shot following hands from left to right, warm harsh overhead light, steam and motion around him, with knife taps and kitchen ambience audible.", "tags": [ "action", "chef", "kitchen", "tracking", "hands" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-022", "model": "happyhorse-1.0", "category": "action", "title": "1965 红色野马敞篷海岸公路", "prompt": "A 1965 cherry-red Mustang convertible drives along a winding California coastal highway at midday, lateral tracking shot from a chase vehicle, hard sunlight, ocean visible in background, with V8 engine roar and wind audible.", "tags": [ "action", "car", "mustang", "tracking", "coast" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-023", "model": "happyhorse-1.0", "category": "action", "title": "越野摩托腾空着陆", "prompt": "A motocross rider launches off a dirt ramp, hangs in the air, lands and accelerates away, wide locked-off static, golden hour backlight, dust cloud at landing, with engine roar and dirt impact audible.", "tags": [ "action", "motocross", "ramp", "static", "golden-hour" ], "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-024", "model": "happyhorse-1.0", "category": "action", "title": "雨后跑道短跑起跑爆发", "prompt": "A sprinter explodes off starting blocks on a rain-soaked track, low-angle wide, locked-off static, overcast flat daylight, motion blur on legs, sharp face, with start beep and spike grip audible.", "tags": [ "action", "sprint", "track", "low-angle", "motion-blur" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-025", "model": "happyhorse-1.0", "category": "action", "title": "街头滑板长坡 Manual 跟随", "prompt": "A street skater executes a manual down a long concrete slope, steadicam follow from behind at ground level, late afternoon, long shadows across pavement, with skateboard wheel hum and trucks audible.", "tags": [ "action", "skateboard", "manual", "steadicam", "afternoon" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-026", "model": "happyhorse-1.0", "category": "action", "title": "乡村道口货运列车飞驰", "prompt": "A freight train passes at full speed through a rural crossing, locked-off static wide, overcast daylight, motion blur on train carriages, static foreground grass, with crossing bell and steel-on-rail rumble audible.", "tags": [ "action", "train", "rural", "static", "motion-blur" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-027", "model": "happyhorse-1.0", "category": "nature", "title": "秋林冠层航拍缓升", "prompt": "Aerial drone pull-back over a dense autumn forest canopy, red and orange foliage stretching to the horizon, overcast diffused light, slow and steady ascent, with drone motor whine and faint wind audible.", "tags": [ "nature", "aerial", "autumn", "forest", "pullback" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-028", "model": "happyhorse-1.0", "category": "nature", "title": "草原积雨云延时", "prompt": "Time-lapse style sequence of storm clouds building over an open prairie, wide locked-off static, dramatic light shifts from gold to grey, tall grass moving in wind, with prairie wind audible throughout.", "tags": [ "nature", "timelapse", "clouds", "prairie", "wind" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-029", "model": "happyhorse-1.0", "category": "nature", "title": "黑色火山岩浪花拍击", "prompt": "A single wave crests and crashes against a black volcanic rock formation, slow dolly in from wide to close-up on the foam, overcast morning light, mist rising, with wave impact and ocean spray audible.", "tags": [ "nature", "wave", "rocks", "dolly", "mist" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-030", "model": "happyhorse-1.0", "category": "nature", "title": "冰川蓝色裂缝航拍横移", "prompt": "Aerial lateral tracking over a glacier surface, deep blue ice crevasses below, snowfields extending to distant peaks, flat cold daylight, slow and deliberate movement, with cold wind and faint ice creak audible.", "tags": [ "nature", "aerial", "glacier", "tracking", "ice" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-031", "model": "happyhorse-1.0", "category": "nature", "title": "蓝调时刻峡谷河流下降航拍", "prompt": "A river winds through a canyon at blue hour, aerial drone descends slowly into the canyon, canyon walls rising on both sides, fading warm light on the water below, with river flow and canyon echo audible.", "tags": [ "nature", "aerial", "canyon", "blue-hour", "river" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-032", "model": "happyhorse-1.0", "category": "cinematic-styles", "title": "东京小巷霓虹孤行", "prompt": "A narrow Tokyo alley at midnight, a lone figure walks away from camera under paper lanterns, steadicam follow from behind, neon pink and cyan reflections in puddles, light rain, with footsteps on wet pavement and distant city hum audible.", "tags": [ "narrative", "tokyo", "alley", "neon", "rain", "steadicam" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-033", "model": "happyhorse-1.0", "category": "cinematic-styles", "title": "高峰地铁站台静止男士", "prompt": "A busy subway platform at rush hour, commuters streaming past a stationary man looking at his phone, locked-off static wide, harsh fluorescent light, motion blur on crowd, with subway rumble and crowd murmur audible.", "tags": [ "narrative", "subway", "crowd", "static", "fluorescent" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-034", "model": "happyhorse-1.0", "category": "cinematic-styles", "title": "伊斯坦布尔晨市铺货", "prompt": "Early morning street market setup in Istanbul, vendors arranging produce stalls, handheld follow drifting between stalls, warm low sunrise light, steam from tea glasses, with tea pouring and vendor chatter audible.", "tags": [ "narrative", "market", "istanbul", "handheld", "sunrise" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-035", "model": "happyhorse-1.0", "category": "cinematic-styles", "title": "脚手架工人远眺晨城", "prompt": "A construction worker looks out from a high scaffold over a waking city at dawn, wide locked-off static, city stretching below, golden hour light from the east, with distant city hum and metal scaffold creak audible.", "tags": [ "narrative", "construction", "dawn", "static", "scaffold" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-036", "model": "happyhorse-1.0", "category": "cinematic-styles", "title": "纽约黄色出租夜雨溅水", "prompt": "A yellow taxi splashes through a puddle on a wet Manhattan intersection at night, locked-off static low-angle wide, neon and traffic light reflections on wet asphalt, light rain, with tire splash and traffic ambience audible.", "tags": [ "narrative", "manhattan", "taxi", "night", "neon", "rain" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-037", "model": "happyhorse-1.0", "category": "product-commercial", "title": "香水瓶黑大理石棚拍旋转", "prompt": "A perfume bottle rotates slowly on a black marble surface, studio rim lighting highlighting glass facets and liquid colour, locked-off static close-up, deep shadow falloff behind, with faint glass turn sound audible.", "tags": [ "product", "perfume", "rotate", "studio", "marble" ], "params": { "aspect": "1:1", "duration": "6s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-038", "model": "happyhorse-1.0", "category": "product-commercial", "title": "运动鞋慢动作落地", "prompt": "A sneaker drops in slow motion onto a white surface and bounces slightly, locked-off static low-angle close-up, hard directional studio light from above right, clean white background, with the rubber sole impact audible.", "tags": [ "product", "sneaker", "slow-motion", "static", "impact" ], "params": { "aspect": "1:1", "duration": "5s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-039", "model": "happyhorse-1.0", "category": "product-commercial", "title": "亚麻桌面咖啡包置入推进", "prompt": "A coffee bag is placed on a raw linen surface by a hand entering from the left, camera slowly dollies in to the product label, warm natural side light, steam suggestion in background, with bag rustle and fabric texture audible.", "tags": [ "product", "coffee", "linen", "dolly", "hand" ], "params": { "aspect": "1:1", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-040", "model": "happyhorse-1.0", "category": "product-commercial", "title": "腕表表盘宏观环绕", "prompt": "A watch face fills the frame, second hand moving, slow lateral orbit around the watch, macro close-up, single cold key light from above, sharp detail on indices and dial texture, with mechanical ticking audible.", "tags": [ "product", "watch", "macro", "orbit", "ticking" ], "params": { "aspect": "1:1", "duration": "6s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-041", "model": "happyhorse-1.0", "category": "product-commercial", "title": "极简白桌笔电开屏推近", "prompt": "A laptop opens on a minimal white desk, screen illuminates, slow dolly in from wide to screen close-up, clean overhead diffused light, no other objects in frame, with hinge mechanism and soft startup chime audible.", "tags": [ "product", "laptop", "minimal", "dolly", "studio" ], "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-042", "model": "happyhorse-1.0", "category": "social-viral", "title": "地中海露台白裙金光", "prompt": "A model in a long white linen dress walks slowly along a whitewashed Mediterranean terrace, steadicam follow from the side, golden hour backlight, fabric moving in sea breeze, with sea breeze and distant waves audible.", "tags": [ "fashion", "mediterranean", "steadicam", "golden-hour", "breeze" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-043", "model": "happyhorse-1.0", "category": "social-viral", "title": "海军蓝西装大理石阶梯下行", "prompt": "A man in a tailored navy suit descends wide marble steps toward camera, slow dolly back, maintaining mid-shot framing, flat overcast daylight, clean architectural background, with leather sole footsteps audible.", "tags": [ "fashion", "suit", "marble", "dolly", "architecture" ], "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-044", "model": "happyhorse-1.0", "category": "social-viral", "title": "腕表表扣特写宏观推进", "prompt": "Close-up on a model's hands fastening a watch clasp, slow push-in from wide hands to macro on clasp, single soft window light from left, warm tones, with the metal clasp click audible.", "tags": [ "fashion", "hands", "watch", "macro", "push-in" ], "params": { "aspect": "1:1", "duration": "5s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-045", "model": "happyhorse-1.0", "category": "social-viral", "title": "皮草大衣雨夜下车霓虹", "prompt": "A woman in a fur-trimmed coat emerges from a car door onto a rain-soaked city street at night, steadicam follow from front, neon light catching coat texture, confident stride, with car door, rain on pavement and heels audible.", "tags": [ "fashion", "rain", "neon", "steadicam", "night" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-046", "model": "happyhorse-1.0", "category": "social-viral", "title": "欧洲阳台晨读悠闲", "prompt": "A man sits reading on a sunlit apartment balcony overlooking a European city, locked-off static mid-shot, warm morning light, coffee cup on railing, slight breeze moving pages, with page turning and distant city ambience audible.", "tags": [ "lifestyle", "balcony", "morning", "static", "european" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-047", "model": "happyhorse-1.0", "category": "cinematic", "title": "日式极简住宅入户穿堂", "prompt": "Slow steadicam push through the front door of a minimal Japanese home into an open living space, warm timber and paper screen interior, low afternoon light cutting across the floor, with sliding door and tatami footstep audible.", "tags": [ "architecture", "japanese", "interior", "steadicam", "afternoon" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-048", "model": "happyhorse-1.0", "category": "cinematic", "title": "悬崖玻璃别墅航拍下降", "prompt": "Aerial drone descends slowly toward a modernist glass house perched on a cliffside above the ocean, late afternoon, long shadows from the roofline, ocean horizon beyond, with drone whine and distant ocean audible.", "tags": [ "architecture", "aerial", "modernist", "cliff", "ocean" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-049", "model": "happyhorse-1.0", "category": "cinematic", "title": "野兽派图书馆横移幽静", "prompt": "A slow lateral dolly moves through a high-ceilinged brutalist concrete library, rows of shelves receding into depth, cold overhead fluorescent light, a single reading lamp in the distance, with quiet hum and faint page turn audible.", "tags": [ "architecture", "brutalist", "library", "dolly", "fluorescent" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-050", "model": "happyhorse-1.0", "category": "cinematic", "title": "走廊穿入阳光内庭", "prompt": "Steadicam push from a dark corridor into a sunlit white courtyard, camera transitions from shadow to bright natural light, potted plants and stone floor, midday overhead sun, with footsteps and faint courtyard birdsong audible.", "tags": [ "architecture", "corridor", "courtyard", "steadicam", "transition" ], "params": { "aspect": "16:9", "duration": "7s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-051", "model": "happyhorse-1.0", "category": "fantasy-scifi", "title": "白色环境服橙色沙漠两月", "prompt": "A lone figure in a white enviro-suit walks across a vast orange desert planet surface, aerial drone pull-back revealing scale, two moons visible in the pale sky, flat cold daylight, with suit servo whir and faint wind audible.", "tags": [ "sci-fi", "desert", "aerial", "worldbuilding", "scale" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-052", "model": "happyhorse-1.0", "category": "fantasy-scifi", "title": "蓝色生物荧光地下枢纽", "prompt": "A futuristic underground transit hub, curved concrete tunnels lit by blue bioluminescent strips, commuters moving as small figures, steadicam push down the central corridor, cold blue light, with low ambient hum and footsteps audible.", "tags": [ "sci-fi", "futuristic", "transit", "steadicam", "bioluminescent" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-053", "model": "happyhorse-1.0", "category": "fantasy-scifi", "title": "古寺夜林荧光藤蔓环绕", "prompt": "An ancient stone temple overgrown with glowing blue bioluminescent vines in a dark jungle, slow lateral orbit around a central monolith, no ambient light source except the vines, with faint jungle insects and rustle audible.", "tags": [ "fantasy", "temple", "jungle", "orbit", "bioluminescent" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-054", "model": "happyhorse-1.0", "category": "fantasy-scifi", "title": "巨型飞船云覆星球入轨", "prompt": "A massive spacecraft enters orbit above a cloud-covered planet, locked-off static wide, spacecraft moving slowly left to right across frame, star field behind, engine glow only light source, with deep engine rumble audible.", "tags": [ "sci-fi", "space", "spacecraft", "static", "engine" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-055", "model": "happyhorse-1.0", "category": "cinematic-styles", "title": "中世纪洪水古城提灯独行", "prompt": "A hooded figure walks through a flooded medieval city at night, lantern in hand, steadicam follow from behind, amber lantern light reflecting on black water surface, deep shadows, with wading water and lantern creak audible.", "tags": [ "fantasy", "medieval", "lantern", "steadicam", "water" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happy-horse-1-0-prompt-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-056", "model": "happyhorse-1.0", "category": "cinematic", "title": "夕阳沙漠旅人剪影航拍环绕", "prompt": "A lone traveler walks slowly across a vast desert at sunset, golden light casting long shadows on rippled sand dunes, wind lifting fine grains into a soft cinematic haze. Low-angle camera follows from behind, gradually circling to reveal the traveler's silhouette against the glowing horizon, with desert wind and footsteps in sand audible.", "tags": [ "cinematic", "desert", "sunset", "silhouette", "tracking" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Atlas Cloud", "url": "https://www.atlascloud.ai/blog/ai-updates/happy-horse1-is-now-on-atlas-cloud", "tier": "reputable-3rd-party" } }, { "id": "hh-057", "model": "happyhorse-1.0", "category": "dialogue-driven", "title": "韩语雨窗咖啡馆独白(原生音频)", "prompt": "A Korean woman in her early 30s sits at a wooden cafe table by a rain-streaked window, speaking directly to camera in Korean with a warm, unhurried tone. Soft ambient cafe noise in the background. Shallow depth of field, warm tungsten lighting, close-up framing, speaking Korean at a natural pace.", "tags": [ "dialogue", "korean", "cafe", "rain", "close-up", "native-audio" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "Imagine.art", "url": "https://www.imagine.art/blogs/happyhorse-1-0-guide", "tier": "reputable-3rd-party" } }, { "id": "hh-058", "model": "happyhorse-1.0", "category": "cinematic", "title": "厨师镜头连贯性测试(压力测试 1)", "prompt": "A male chef, mid-40s, silver-streaked hair, thin-framed glasses. He cuts daikon with one clean stroke. Tends a clay pot on the iron stove.", "tags": [ "chef", "consistency", "cinematic", "stress-test", "20-word-rule" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Alibaba's New AI Video Model Tested (YouTube)", "url": "https://www.youtube.com/watch?v=_Y4siSt-dG8", "tier": "reputable-3rd-party" } }, { "id": "hh-059", "model": "happyhorse-1.0", "category": "action", "title": "BMX 角斗士运动物理测试", "prompt": "A BMX rider in gladiator armor, no helmet, face always visible, rides a red and gold BMX through the Roman Colosseum during a packed gladiatorial event.", "tags": [ "action", "bmx", "gladiator", "colosseum", "physics" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Alibaba's New AI Video Model Tested (YouTube)", "url": "https://www.youtube.com/watch?v=_Y4siSt-dG8", "tier": "reputable-3rd-party" } }, { "id": "hh-060", "model": "happyhorse-1.0", "category": "cinematic", "title": "东南亚街道无人机运镜", "prompt": "Busy Southeast Asian street packed with motorbikes and traffic, lined with French colonial buildings. Drone starts at street level, speeds through traffic, pulls up sharply skimming the building facade.", "tags": [ "drone", "fpv", "street", "southeast-asia", "tracking" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Alibaba's New AI Video Model Tested (YouTube)", "url": "https://www.youtube.com/watch?v=_Y4siSt-dG8", "tier": "reputable-3rd-party" } }, { "id": "hh-061", "model": "happyhorse-1.0", "category": "dialogue-driven", "title": "街头小吃口型同步测试", "prompt": "The man holds hot street food, takes a bite, smiles slightly and says, \"Yo, this is actually good.\" Natural head movement, blinking, and clean lip sync throughout.", "tags": [ "dialogue", "lip-sync", "street-food", "i2v", "natural" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "Alibaba's New AI Video Model Tested (YouTube)", "url": "https://www.youtube.com/watch?v=_Y4siSt-dG8", "tier": "reputable-3rd-party" } }, { "id": "hh-062", "model": "happyhorse-1.0", "category": "video-edit", "title": "Edit Video 模式:换衬衫颜色", "prompt": "Change the man's shirt color from black to white.", "tags": [ "video-edit", "color-change", "surgical-edit", "minimal-prompt" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Alibaba's New AI Video Model Tested (YouTube)", "url": "https://www.youtube.com/watch?v=_Y4siSt-dG8", "tier": "reputable-3rd-party" } }, { "id": "hh-063", "model": "happyhorse-1.0", "category": "action", "title": "卡车追逐场景(20 词内)", "prompt": "a tracking shot of the man slowly walking towards the truck, a thug hops out of the car holding a shotgun. He takes a shot, and the man dodges.", "tags": [ "tracking", "action", "chase", "20-word-rule" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Happy Horse: The Seedance Killer? (YouTube)", "url": "https://www.youtube.com/watch?v=JVWTwKrus9c", "tier": "reputable-3rd-party" } }, { "id": "hh-064", "model": "happyhorse-1.0", "category": "comedy", "title": "西装男逃避罚单(极简短 prompt)", "prompt": "a man in a blue business suit running away from a jaywalking ticket", "tags": [ "comedy", "absurd", "minimal-prompt", "20-word-rule" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Happy Horse: The Seedance Killer? (YouTube)", "url": "https://www.youtube.com/watch?v=JVWTwKrus9c", "tier": "reputable-3rd-party" } }, { "id": "hh-065", "model": "happyhorse-1.0", "category": "cinematic", "title": "霓虹雨城单镜头电影感", "prompt": "a cinematic close-up of a woman speaking in a rainy neon city with ambient sound and dialogue", "tags": [ "cinematic", "close-up", "rain", "neon", "dialogue", "minimal-prompt" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Pollo AI + Happy Horse (YouTube)", "url": "https://www.youtube.com/watch?v=U4M_bEJIGPI", "tier": "reputable-3rd-party" } }, { "id": "hh-066", "model": "happyhorse-1.0", "category": "portrait", "title": "GPT Image 2 角色生成 — 2D 动画肖像", "prompt": "A high-quality 2D stylized digital animation portrait, strictly flat 2D illustration aesthetic with no 3D depth or CGI realism. The character is a young girl with an extremely sharp, geometric jawline and angular facial features defined by straight lines rather than curves. She has wild, wind-swept vibrant orange hair (#E67E22) with painterly textures and fair freckled skin. She looks directly forward with wide, intense blue eyes. She is wearing oversized octagonal thin-frame glasses with complex, multi-faceted gem-cut crystal lenses that refract a brilliant iridescent rainbow spectrum (cyan, magenta, yellow) across the eyes. Clothing: A simple dark charcoal grey hoodie. Background: Solid, seamless white background for a clean character isolate. Art Style: Hard-edged 2D line work, graphic cel-shading, and bold geometric shapes. The lighting is soft and flat with minimal angular graphic shadows under the chin. Professional character concept art style, symmetrical composition.", "tags": [ "character-design", "2d-illustration", "portrait", "character-sheet", "gpt-image-2" ], "params": { "aspect": "1:1", "duration": "n/a (image)" }, "source": { "name": "Happy Horse Better Than Seedance IF You Do This (Google Doc, YouTube)", "url": "https://www.youtube.com/watch?v=4hhpIyznnaw", "tier": "reputable-3rd-party" } }, { "id": "hh-067", "model": "happyhorse-1.0", "category": "action", "title": "高速城市骑行追逐(5 镜头时间码 + AUDIO 区块)", "prompt": "CHARACTER CYCLIST, intense focused expression, athletic posture. Bright sun urban highway and city streets, multi-lane traffic, overpasses, suspension bridge, red-marked tram tracks, clear blue sky.\n\nRIDE 1 (0:00-0:02): Low-angle rear tracking. CHARACTER pedals hard, lane changes, weaving between cars, sharp lean into turns, fast motion, real-time, motion-blur background.\n\nRIDE 2 (0:02-0:02.5): Quick close-up of CHARACTER's face, intense focus forward. Micro-cut, blurred background, real-time.\n\nRIDE 3 (0:02.5-0:04): Low-angle front tracking. CHARACTER zigzags through traffic, quick handlebar flicks, swerves past bumpers, camera shake, fast tracking.\n\nRIDE 4 (0:04-0:06): 3/4 front-side tracking. Slalom across lanes, close passes to cars, aggressive lean, camera swings side-to-side following weave.\n\nRIDE 5 (0:06-0:08): Low-angle rear tracking. CHARACTER overtakes dark blue car, cuts in front through narrow gap, speed ramp, city streaking behind.\n\nAUDIO (ref): Heavy breathing, pedaling, traffic whoosh, wind rush, tire friction, quick tire squeal.", "tags": [ "multi-shot", "timeline-prompting", "cycling", "tracking", "audio-block", "speed-ramp" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Happy Horse Better Than Seedance IF You Do This (Google Doc, YouTube)", "url": "https://www.youtube.com/watch?v=4hhpIyznnaw", "tier": "reputable-3rd-party" } }, { "id": "hh-068", "model": "happyhorse-1.0", "category": "product-commercial", "title": "运动鞋电影广告(Prompt Lab 自动展开)", "prompt": "Turn this sports shoe into a premium cinematic advertisement, dynamic camera rotation, dramatic lighting, dust particles, athlete energy, fast push-in, high-contrast product reveal.", "tags": [ "product", "shoe", "advertisement", "push-in", "prompt-lab", "cinematic" ], "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "New #1 Ranked AI Video Generator Is Here (YouTube)", "url": "https://www.youtube.com/watch?v=q4rhf7msTkA", "tier": "reputable-3rd-party" } }, { "id": "hh-069", "model": "happyhorse-1.0", "category": "fantasy-scifi", "title": "R2V 城堡王子电影跟拍(4 张参考图)", "prompt": "The prince walking forward through the castle grounds, cloak moving in the wind, guards in the distance, banners shifting, sunlight cutting through the clouds, and the camera following with a smooth cinematic track.", "tags": [ "r2v", "fantasy", "castle", "prince", "cinematic-tracking", "multi-reference" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Alibaba HappyHorse Created My Childhood Game in Real Life (YouTube)", "url": "https://www.youtube.com/watch?v=MIeRMIT3nYA", "tier": "reputable-3rd-party" } }, { "id": "hh-070", "model": "happyhorse-1.0", "category": "fantasy-scifi", "title": "T2V 中世纪庭院日出建立镜头", "prompt": "Giant stone towers, guards, villagers, horses, market carts, banners, smoke, golden sunrise light, and a slow cinematic tracking camera moving into the environment.", "tags": [ "t2v", "medieval", "establishing-shot", "sunrise", "tracking", "world-building" ], "params": { "aspect": "16:9", "duration": "15s" }, "source": { "name": "Alibaba HappyHorse Created My Childhood Game in Real Life (YouTube)", "url": "https://www.youtube.com/watch?v=MIeRMIT3nYA", "tier": "reputable-3rd-party" } }, { "id": "hh-071", "model": "happyhorse-1.0", "category": "quiet-moments", "title": "亚裔女孩窗边沉思情感短片", "prompt": "An Asian girl sitting by a window in the evening natural light, holding a drink, quietly looking out the window, occasionally glancing down and then back out, relaxed and thoughtful. The camera work also needs to be slow and deliberate, creating an atmospheric short film with continuous, natural visuals, delicate movements, and consistent emotions.", "tags": [ "quiet", "emotional", "window", "evening", "continuous-action", "slow-camera" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "Stop Wasting Money on AI Videos (LitMedia AI, YouTube)", "url": "https://www.youtube.com/watch?v=L7hUBvxYEo8", "tier": "reputable-3rd-party" } }, { "id": "hh-072", "model": "happyhorse-1.0", "category": "dialogue-driven", "title": "三语 Lip-Sync 测试(英 / 普通话 / 西班牙语)", "prompt": "You shouldn't have come here tonight. It was a mistake.\nNǐ yǐwéi nǐ neng táo diào ma? Bié tài tiānzhēn le.\nAhora, el silencio es lo único que te queda.", "tags": [ "lip-sync", "multilingual", "english", "mandarin", "spanish", "dialogue" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Seedance 2.0 vs HappyHorse vs Kling 3.0 — 15 Prompts (YouTube)", "url": "https://www.youtube.com/watch?v=UN7bJDvqclg", "tier": "reputable-3rd-party" } }, { "id": "hh-073", "model": "happyhorse-1.0", "category": "social-viral", "title": "TopView AI Storyboard 极简引用", "prompt": "generate a video based on a storyboard at image one", "tags": [ "storyboard", "minimal-prompt", "i2v", "social-viral", "topview" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "Seedance 2.0 vs HappyHorse vs Kling 3.0 — 15 Prompts (YouTube)", "url": "https://www.youtube.com/watch?v=UN7bJDvqclg", "tier": "reputable-3rd-party" } }, { "id": "hh-074", "model": "happyhorse-1.0", "category": "experimental", "title": "连续缩放压力测试(卫星→TikTok)", "prompt": "A satellite view of planet Earth, diving down into a drone view of New York City, pushing through an office building window, and finally ending on a tight close-up of a person scrolling through TikTok.", "tags": [ "experimental", "continuous-zoom", "scale-test", "drone", "stress-test", "failure-mode" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Alibaba's New AI Video Model is Ranked #1... But is it Actually Good? (YouTube)", "url": "https://www.youtube.com/watch?v=MO6UGe4w0BM", "tier": "reputable-3rd-party" } }, { "id": "hh-075", "model": "happyhorse-1.0", "category": "fantasy-scifi", "title": "公主与火龙多阶段叙事(失败边界)", "prompt": "A princess in a glittery white dress is running from a massive dragon in a dense forest. The dragon breathes fire, the ground ignites, and the princess has to jump onto drifting debris to cross a river while the dragon roars in frustration.", "tags": [ "fantasy", "sequential-narrative", "stress-test", "dragon", "physics", "failure-mode" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Alibaba's New AI Video Model is Ranked #1... But is it Actually Good? (YouTube)", "url": "https://www.youtube.com/watch?v=MO6UGe4w0BM", "tier": "reputable-3rd-party" } }, { "id": "hh-076", "model": "happyhorse-1.0", "category": "social-viral", "title": "UGC 抗痘霜青少年代言(带原生音频)", "prompt": "A teenage girl in her bedroom, natural lighting, subtle acne texture on her skin. She applies anti-acne cream and says, \"Two weeks in and my skin is finally calming down.\" Casual, authentic UGC style. Natural hand movement, realistic bedroom clutter in background.\nAUDIO: Her voice, soft room tone, no music.", "tags": [ "ugc", "skincare", "dialogue", "audio-block", "authentic", "social-viral" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "I Tested HappyHorse 1.0 Inside Pollo AI (YouTube)", "url": "https://www.youtube.com/watch?v=OPElYdtdgkQ", "tier": "reputable-3rd-party" } }, { "id": "hh-077", "model": "happyhorse-1.0", "category": "product-commercial", "title": "GPT Image 2 香水瓶高端产品图", "prompt": "Transparent blue glass bottle on white marble, soft golden light, no label, no text. Premium perfume product shot.", "tags": [ "product", "perfume", "gpt-image-2", "minimal-prompt", "premium" ], "params": { "aspect": "1:1", "duration": "n/a (image)" }, "source": { "name": "I Tested HappyHorse 1.0 Inside Pollo AI (YouTube)", "url": "https://www.youtube.com/watch?v=OPElYdtdgkQ", "tier": "reputable-3rd-party" } }, { "id": "hh-078", "model": "happyhorse-1.0", "category": "product-commercial", "title": "Eclora 香水广告对白", "prompt": "Eclora feels fresh, soft, and unforgettable.", "tags": [ "product", "perfume", "dialogue", "minimal-prompt", "advertisement" ], "params": { "aspect": "9:16", "duration": "5s" }, "source": { "name": "I Tested HappyHorse 1.0 Inside Pollo AI (YouTube)", "url": "https://www.youtube.com/watch?v=OPElYdtdgkQ", "tier": "reputable-3rd-party" } }, { "id": "hh-079", "model": "happyhorse-1.0", "category": "product-commercial", "title": "电池能量饮料 UGC(带原生音频 + 道具声)", "prompt": "A guy picks up the energy drink can, pulls it away from the camera, brings it closer, takes a sip, places it back on the desk. He says, \"This looks like a battery because it basically recharges my day.\" Natural beard, realistic skin texture, UGC casual vibe.\nAUDIO: His voice, room tone, sound of can being placed back on desk.", "tags": [ "ugc", "product", "energy-drink", "dialogue", "audio-block", "foley" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "I Tested HappyHorse 1.0 Inside Pollo AI (YouTube)", "url": "https://www.youtube.com/watch?v=OPElYdtdgkQ", "tier": "reputable-3rd-party" } }, { "id": "hh-080", "model": "happyhorse-1.0", "category": "product-commercial", "title": "电池能量罐 GPT Image 2 道具设计", "prompt": "An energy drink can with copper on the top and matte black on the bottom, inspired by an alkaline battery visual. Pull tab, aluminum body, condensation. No logos, no text.", "tags": [ "product-design", "gpt-image-2", "energy-drink", "minimal-prompt", "concept" ], "params": { "aspect": "1:1", "duration": "n/a (image)" }, "source": { "name": "I Tested HappyHorse 1.0 Inside Pollo AI (YouTube)", "url": "https://www.youtube.com/watch?v=OPElYdtdgkQ", "tier": "reputable-3rd-party" } }, { "id": "hh-081", "model": "happyhorse-1.0", "category": "social-viral", "title": "WWE 故事板 GPT Image 2(16 帧)", "prompt": "A cinematic women's championship wrestling match storyboard featuring intense WWE style combat, emotional moments, dramatic reversals, powerful finishing moves, and massive arena energy. 16 frames with continuous visual progression between every shot.", "tags": [ "storyboard", "wwe", "gpt-image-2", "wrestling", "16-frames", "social-viral" ], "params": { "aspect": "16:9", "duration": "n/a (image)" }, "source": { "name": "Happy Horse AI + GPT Image 2 Combo in Pollo AI (YouTube)", "url": "https://www.youtube.com/watch?v=RSEnVScjkCA", "tier": "reputable-3rd-party" } }, { "id": "hh-082", "model": "happyhorse-1.0", "category": "action", "title": "WWE 冠军决斗(Storyboard → 视频)", "prompt": "Create a cinematic WWE style championship fight sequence that strictly follows the storyboard progression with realistic movement, dramatic reversals, crowd energy, and a massive final finisher.", "tags": [ "action", "wwe", "storyboard-following", "wrestling", "crowd", "long-video" ], "params": { "aspect": "16:9", "duration": "15s" }, "source": { "name": "Happy Horse AI + GPT Image 2 Combo in Pollo AI (YouTube)", "url": "https://www.youtube.com/watch?v=RSEnVScjkCA", "tier": "reputable-3rd-party" } }, { "id": "hh-083", "model": "happyhorse-1.0", "category": "social-viral", "title": "地堡建造 10x 延时(Storyboard 序列)", "prompt": "Create a brand new realistic live-action cinematic video that strictly follows the storyboard sequence while showing a man manually constructing an underground bunker in his backyard in an ultra fast 10 times time-lapse style.", "tags": [ "time-lapse", "diy", "bunker", "storyboard-following", "social-viral", "long-form" ], "params": { "aspect": "9:16", "duration": "15s" }, "source": { "name": "Happy Horse AI + GPT Image 2 Combo in Pollo AI (YouTube)", "url": "https://www.youtube.com/watch?v=RSEnVScjkCA", "tier": "reputable-3rd-party" } }, { "id": "hh-084", "model": "happyhorse-1.0", "category": "fantasy-scifi", "title": "动物融合变形(Storyboard 连续)", "prompt": "Create a cinematic fusion transformation video that strictly follows the attached storyboard sequence while showing both creatures evolving from an intense face-off to collision and finally into a realistic hybrid creature with seamless transformation continuity.", "tags": [ "transformation", "fantasy", "fusion", "storyboard-following", "creature", "vfx" ], "params": { "aspect": "16:9", "duration": "15s" }, "source": { "name": "Happy Horse AI + GPT Image 2 Combo in Pollo AI (YouTube)", "url": "https://www.youtube.com/watch?v=RSEnVScjkCA", "tier": "reputable-3rd-party" } }, { "id": "hh-085", "model": "happyhorse-1.0", "category": "social-viral", "title": "世界旅行 Vlog 分段生成(场景 1-5)", "prompt": "Generate scenes one to five from the storyboard reference. Cinematic selfie-style journey, female influencer recording in front camera POV, consistent character appearance, realistic travel moments.", "tags": [ "travel", "vlog", "selfie-pov", "storyboard-segmenting", "influencer", "long-form" ], "params": { "aspect": "9:16", "duration": "15s" }, "source": { "name": "Happy Horse AI + GPT Image 2 Combo in Pollo AI (YouTube)", "url": "https://www.youtube.com/watch?v=RSEnVScjkCA", "tier": "reputable-3rd-party" } }, { "id": "hh-086", "model": "happyhorse-1.0", "category": "cinematic-styles", "title": "黑伞悬崖变形(五要素摄像机优先)", "prompt": "Subject: a black umbrella. Action: thrown and then changes. Environment: large cliffs plus valleys. Camera: orbital camera plus FPV drone. Atmosphere: golden atmosphere plus volumetric light.", "tags": [ "transformation", "five-element", "orbital", "fpv-drone", "golden-hour", "volumetric" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "720p" }, "source": { "name": "HappyHorse Tips for Cinematic Camera Movement in TopMediai (YouTube)", "url": "https://www.youtube.com/watch?v=SkjumXCTnbQ", "tier": "reputable-3rd-party" } }, { "id": "hh-087", "model": "happyhorse-1.0", "category": "action", "title": "JXP 黑暗主角拔武士刀(3 张参考图)", "prompt": "Dark protagonist unsheathes katana in slow motion. Cinematic action film sequence, ultra fluid motion.", "tags": [ "action", "katana", "slow-motion", "reference-mode", "jxp", "minimal-prompt" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "720p" }, "source": { "name": "HappyHorse 1.0 on JXP (YouTube)", "url": "https://www.youtube.com/watch?v=aKjV3iL0Nqo", "tier": "reputable-3rd-party" } }, { "id": "hh-088", "model": "happyhorse-1.0", "category": "experimental", "title": "Reference 模式错误示例(双角色开场帧)", "prompt": "on a New York street, image one sees image two", "tags": [ "experimental", "failure-mode", "reference", "multi-character", "minimal-prompt" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Happy Horse: The Seedance Killer? (YouTube)", "url": "https://www.youtube.com/watch?v=JVWTwKrus9c", "tier": "reputable-3rd-party" } }, { "id": "hh-089", "model": "happyhorse-1.0", "category": "action", "title": "ATV 沙地追逐多镜头(Freepik Prompt Editor)", "prompt": "An intense chase scene on an ATV. Sequence of events: ATV enters frame at speed; close-up of driver's face, intense focus; wide shot of terrain, dust clouds; obstacle appears; dramatic evasion maneuver. Camera: Following tracking shot, low-angle, motion blur. AUDIO: Engine roar, gravel spray, wind rush.", "tags": [ "action", "atv", "chase", "multi-shot", "freepik", "audio-block", "tracking" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "The New #1 AI Video Tool Just Dropped, Happy Horse (YouTube)", "url": "https://www.youtube.com/watch?v=gSLMDB4Uj28", "tier": "reputable-3rd-party" } }, { "id": "hh-090", "model": "happyhorse-1.0", "category": "experimental", "title": "Freepik Prompt Editor 自动展开指令", "prompt": "Make it more detailed, cinematic multi-shot, specifically for happy horse 1.0, and best prompt adherence strategies.", "tags": [ "prompt-editor", "freepik", "meta-prompt", "multi-shot", "experimental" ], "params": { "aspect": "16:9", "duration": "n/a" }, "source": { "name": "The New #1 AI Video Tool Just Dropped, Happy Horse (YouTube)", "url": "https://www.youtube.com/watch?v=gSLMDB4Uj28", "tier": "reputable-3rd-party" } }, { "id": "hg-001", "model": "higgsfield-soul", "category": "cinematic", "title": "Higgsfield DoP 模板:Push In", "prompt": "Slow push in. Medium close-up of a woman with dark wet hair under a flickering streetlamp at night. Her eyes track something off-frame to the left. Rain falls steadily, lit only by the lamp.", "tags": [ "dop-template", "push-in", "noir", "short" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Higgsfield", "url": "https://higgsfield.ai/", "tier": "official" }, "notes": "Higgsfield 反常识:短 prompt 优于长 prompt。DoP 模板直接给运镜动词。" }, { "id": "hg-002", "model": "higgsfield-soul", "category": "social-viral", "title": "Higgsfield DoP 模板:Whip Pan", "prompt": "Whip pan from left to right. A skater lands a trick on wet pavement at night. Neon signs streak in the motion blur. Cut on the landing impact.", "tags": [ "dop-template", "whip-pan", "viral", "short" ], "params": { "aspect": "9:16", "duration": "4s" }, "source": { "name": "Higgsfield", "url": "https://higgsfield.ai/skills", "tier": "official" } }, { "id": "hg-003", "model": "higgsfield-soul", "category": "portrait", "title": "Higgsfield Soul ID 角色保持样板", "prompt": "[Soul ID: my-character] walks into a sunlit kitchen wearing a white linen shirt, opens the fridge, takes out a bottle of water, turns to camera with a slight smile. Warm morning light. Medium shot, eye level.", "tags": [ "soul-id", "character-consistency", "everyday" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "Higgsfield Soul", "url": "https://higgsfield.ai/soul-intro", "tier": "official" }, "notes": "Soul ID 是 Higgsfield 独家:训练一次后,角色脸在所有场景保持一致。[Soul ID: name] 占位符替换" }, { "id": "hg-004", "model": "higgsfield-soul", "category": "cinematic", "title": "Higgsfield DoP 模板:Crash Zoom", "prompt": "Crash zoom from wide establishing to extreme close-up on the protagonist eye, the moment of realization frozen in his iris where a tiny silhouette approaches from far behind, cold cyan rim light with dim ambient fill, shallow focus, 24fps, single take with consistent character ID, no morphing of background plates.", "tags": [ "dop-template", "crash-zoom", "thriller", "short", "soul-id" ], "params": { "aspect": "21:9", "duration": "4s" }, "source": { "name": "Higgsfield · DoP Skills (Crash Zoom)", "url": "https://higgsfield.ai/skills", "tier": "official" } }, { "id": "hg-005", "model": "higgsfield-soul", "category": "viral-transform", "title": "Higgsfield Soul + 变身爆款", "prompt": "[Soul ID: my-character] stands in casual street clothes. Lightning flash. In the next frame she wears full cyberpunk samurai armor with glowing katana. Cool neon palette. Vertical short.", "tags": [ "soul-id", "transformation", "cyberpunk", "viral" ], "params": { "aspect": "9:16", "duration": "5s" }, "source": { "name": "Higgsfield", "url": "https://higgsfield.ai/blog/ai-short-film-youtube-guide", "tier": "official" } }, { "id": "hg-006", "model": "higgsfield-soul", "category": "music-video", "title": "Higgsfield 60s 智能分镜 MV", "prompt": "60-second music video. [Soul ID: singer]. Scene 1: rooftop sunset, singer performs to camera, golden hour. Scene 2: car interior at night, singer drives through neon-lit tunnel, intercut close-ups. Scene 3: empty stage with single spotlight, singer holds final note, slow pull-back to reveal massive empty arena. Energetic synth-pop aesthetic throughout, cuts on beat.", "tags": [ "music-video", "multi-scene", "soul-id", "ai-director" ], "params": { "aspect": "9:16", "duration": "60s" }, "source": { "name": "Higgsfield AI Director", "url": "https://higgsfield.ai/", "tier": "official" }, "notes": "Higgsfield AI Director 强项:多场景自动分镜 + Soul ID 角色保持" }, { "id": "hg-007", "model": "higgsfield-soul", "category": "product-commercial", "title": "Higgsfield 化妆品 viral 短片", "prompt": "Tabletop hero shot. Close-up of a frosted glass lipstick container on a marble surface. Soft pink and gold lighting. The lid pops off with a satisfying click, the lipstick slowly extends. Camera orbits halfway around. Particle sparkles drift across the frame. Aspirational beauty commercial tone.", "tags": [ "product", "beauty", "viral", "short" ], "params": { "aspect": "9:16", "duration": "6s" }, "source": { "name": "Higgsfield", "url": "https://blog.segmind.com/higgsfield-ai-prompt-guide-video-creation/", "tier": "reputable-3rd-party" } }, { "id": "hg-008", "model": "higgsfield-soul", "category": "cinematic", "title": "Higgsfield 90s 短片完整结构", "prompt": "90-second cinematic short. [Soul ID: detective]. Opening: rainy noir alleyway, detective lights a cigarette, voiceover begins. Middle: flashback intercut — daytime cafe meeting with mystery client. Climax: chase through wet streets, jump cuts, motion blur. Ending: detective walks away into the rain, camera pulls back to high crane shot, title card fades in. Cold noir palette throughout with selective warm accents on key emotional beats.", "tags": [ "short-film", "noir", "soul-id", "ai-director" ], "params": { "aspect": "21:9", "duration": "90s" }, "source": { "name": "Higgsfield AI Director", "url": "https://higgsfield.ai/blog/ai-short-film-youtube-guide", "tier": "official" } }, { "id": "hy-001", "model": "hunyuan-video", "category": "cinematic", "title": "Hunyuan 跨维度:虚拟角色融入真实场景", "prompt": "A stylized 3D anime character with long silver hair walks naturally through a real photographic shopping street, lighting and reflections on her hair match the real-world overcast sky perfectly. Medium tracking shot. Seamless cross-dimensional integration.", "tags": [ "cross-dimensional", "hybrid", "tracking" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Tencent Hunyuan", "url": "https://github.com/Tencent-Hunyuan/HunyuanVideo", "tier": "official-platform" }, "notes": "Hunyuan 独有的「跨维度生成」能力:虚拟角色自然融入真实场景" }, { "id": "hy-002", "model": "hunyuan-video", "category": "open-source", "title": "本地部署测试 prompt(Master 模式)", "prompt": "A 1990s Hong Kong neon-lit market street at dusk, vendors arranging fresh fish on ice, steam rising from food stalls, a young woman in a yellow raincoat browses with curiosity. Handheld camera with slight shake, warm tungsten light against cool dusk sky, 35mm film grain, Wong Kar-wai aesthetic.", "tags": [ "hk-cinema", "atmospheric", "open-source", "test" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Hunyuan", "url": "https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5/blob/main/assets/HunyuanVideo_1_5_Prompt_Handbook_EN.md", "tier": "official-platform" }, "notes": "适合 Master 模式的复杂场景描述,开源本地部署测试用" }, { "id": "hy-003", "model": "hunyuan-video", "category": "image-to-video", "title": "Hunyuan I2V 只描述运动", "prompt": "The wind picks up; the model's hair lifts and floats around her face, her dress flutters at the hem, she slowly turns her head to glance at the camera with a faint smile. Slight forward camera drift.", "tags": [ "image-to-video", "motion-only", "hair", "fabric" ], "params": { "aspect": "9:16", "duration": "5s" }, "source": { "name": "ComfyUI Hunyuan", "url": "https://docs.comfy.org/tutorials/video/hunyuan/hunyuan-video", "tier": "official" }, "notes": "图生视频只描述运动,不重描述图中的视觉内容(Hunyuan 重视首帧)" }, { "id": "hy-004", "model": "hunyuan-video", "category": "open-source", "title": "简单场景(Normal 模式测试基础能力)", "prompt": "A small fluffy white kitten chases a feather toy across a sunlit hardwood floor, ears perked up, paws batting playfully. Side-angle medium shot.", "tags": [ "pet", "simple", "test", "open-source" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Hunyuan", "url": "https://github.com/Tencent-Hunyuan/HunyuanVideo", "tier": "official-platform" }, "notes": "Normal 模式适合简单场景,先用基础提示测试模型" }, { "id": "hy-005", "model": "hunyuan-video", "title": "古代水墨山水画动起来", "prompt": "A traditional Chinese ink-wash landscape painting comes alive on flat, laid-out Xuan paper with a subtle, fibrous texture. Initially, the painting appears still: several steep distant mountains stand faintly visible amidst clouds and mist, a small wooden boat rests on a wide, tranquil river. Then, soft diluted grey tones begin to billow around the mountains, creating layers of atmospheric mist that drift slowly. The boat slides gently downstream as ink ripples extend outward from its bow. A heron rendered in a few deft strokes lifts off from the reeds, its wings unfolding in fluid ink lines. Subtle motion throughout, minimalist composition. The mood is elegant, poetic, with cultural depth — the ink itself seems to breathe.", "tags": [ "ink-wash", "chinese", "animation", "minimal" ], "category": "fantasy-scifi", "params": { "aspect": "21:9", "duration": "8s" }, "source": { "name": "Tencent Hunyuan Video 1.5", "url": "https://huggingface.co/tencent/HunyuanVideo-1.5", "tier": "official-platform" } }, { "id": "hy-006", "model": "hunyuan-video", "category": "action", "title": "武侠飞檐走壁", "prompt": "A Chinese wuxia swordsman in flowing dark robes leaps across rooftops at twilight, robes billow behind him, sword catches the last light. Tracking side-shot, dust trails behind each step, cinematic ink-meets-cinema aesthetic.", "tags": [ "wuxia", "action", "tracking", "chinese" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Hunyuan", "url": "https://huggingface.co/tencent/HunyuanVideo-1.5", "tier": "official-platform" } }, { "id": "hy-007", "model": "hunyuan-video", "title": "工坊老工匠手部特写", "prompt": "A macro shot focuses on an elderly Chinese craftsman's weathered hands as they carve intricate patterns into a wooden block, rendered in a cinematic photography realistic style. The hands are positioned diagonally across the frame, deep grooves and age spots visible on the skin. Initially, the chisel is held with surgical steadiness above the wood. Then, with breathtaking precision and in extreme slow motion, the blade descends, and a curl of pale wood shavings peels away in a single continuous spiral. Next, dust motes begin to drift through a single beam of warm window light that crosses the workbench at a low angle. The camera holds steady on the hands and the emerging pattern, capturing each careful pass of the chisel, the fine grain of the wood, and the quiet dignity of the craft.", "tags": [ "craft", "hands", "documentary", "close-up" ], "category": "portrait", "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Tencent Hunyuan Video 1.5", "url": "https://huggingface.co/tencent/HunyuanVideo-1.5", "tier": "official-platform" } }, { "id": "hy-008", "model": "hunyuan-video", "category": "nature", "title": "竹林雪雾", "prompt": "A snowy bamboo forest at dawn, mist drifts between tall green bamboo, gentle snowfall, a small bird shakes snow off a branch. Slow pan from left to right. Cool blue palette with warm dawn glow on bamboo tops. Tranquil ambient mood.", "tags": [ "nature", "bamboo", "snow", "tranquil" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Replicate Hunyuan", "url": "https://replicate.com/tencent/hunyuan-video/examples", "tier": "official-platform" } }, { "id": "hy-009", "model": "hunyuan-video", "category": "cinematic", "title": "复古胶片质感舞会", "prompt": "A 1920s art deco ballroom, couples slow dancing under a glittering chandelier, women in shimmering flapper dresses, men in tuxedos. Slow circular tracking shot. Warm amber tungsten light, 35mm film grain, slight gate weave.", "tags": [ "vintage", "1920s", "ballroom", "tracking" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Replicate", "url": "https://replicate.com/tencent/hunyuan-video/examples", "tier": "official-platform" } }, { "id": "hy-010", "model": "hunyuan-video", "category": "creative", "title": "梵高星空风格动画", "prompt": "A starry night sky in Van Gogh's swirling brushstroke style, stars twinkle and the swirl pattern animates slowly, a small village below with warm yellow window lights. Painterly textures preserved, oil-on-canvas feel.", "tags": [ "art-style", "van-gogh", "painterly", "starry" ], "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "fal.ai", "url": "https://fal.ai/models/fal-ai/hunyuan-video", "tier": "official-platform" } }, { "id": "hy-011", "model": "hunyuan-video", "category": "image-to-video", "title": "Hunyuan I2V 风景动画(湖面涟漪)", "prompt": "Gentle ripples spread across the lake surface, fog slowly drifts to the right, tall pine trees sway subtly with a breeze. Camera holds completely still. Atmospheric ambient feel.", "tags": [ "image-to-video", "lake", "atmospheric", "static-cam" ], "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "CrePal", "url": "https://crepal.ai/blog/aivideo/aivideo-hunyuan-image-to-video-tutorial/", "tier": "reputable-3rd-party" } }, { "id": "hy-012", "model": "hunyuan-video", "title": "LoRA 风格化训练样本(赛博朋克水墨)", "prompt": "A cyberpunk samurai in glowing neon armor walks through a rain-soaked Tokyo alleyway at night. The scene is rendered in a fusion style: the structural lines of the figure, the buildings, and the rain itself flow in traditional Chinese ink-wash brushstrokes — varying shades from jet-black to smoky grey — while highlights and reflections burn in saturated cyberpunk pinks, purples, and electric blues. The wet asphalt below catches shimmering neon reflections, and a complex web of cables crisscrosses overhead. As the samurai moves forward, trails of smoke and rain dissolve into ink particles that drift and reform in the air behind him. The camera tracks slowly in parallel, holding a medium full shot. Atmospheric haze fills the alley; distant flickering advertisements paint the mist in shifting colors.", "tags": [ "cyberpunk", "ink-wash", "fusion", "lora-test" ], "category": "open-source", "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Tencent Hunyuan Video 1.5", "url": "https://huggingface.co/tencent/HunyuanVideo-1.5", "tier": "official-platform" } }, { "id": "jm-001", "model": "jimeng-3", "category": "chinese-native", "title": "即梦中文样板:江南水乡晨景", "prompt": "清晨的江南水乡古镇,薄雾笼罩青石板路,乌篷船缓缓驶过石拱桥,岸边垂柳轻摇,一位身穿淡蓝色棉麻长裙的女子撑着油纸伞独自走过。镜头中景平稳跟拍,暖黄色晨光透过晨雾,柔和电影质感,淡淡国风韵味。", "tags": [ "chinese", "jiangnan", "atmospheric", "tracking" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" }, "notes": "即梦中文理解最强,复杂中式场景用纯中文 prompt 效果最佳" }, { "id": "jm-002", "model": "jimeng-3", "category": "chinese-native", "title": "即梦古风:汉服女子起舞", "prompt": "一位身穿绛红色汉服的年轻女子在古典园林中翩翩起舞,长袖飘逸,发簪上的玉坠随动作轻摆,背景是粉色海棠盛开。镜头缓慢环绕,黄昏柔光,杏花瓣随风飘落。整体氛围典雅唯美,水墨意境。", "tags": [ "chinese", "hanfu", "dance", "orbit" ], "params": { "aspect": "9:16", "duration": "10s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "jm-003", "model": "jimeng-3", "category": "cinematic", "title": "即梦电影感:地铁站离别", "prompt": "黄昏的地铁站台,一对年轻情侣紧紧拥抱告别,列车即将进站,远处可见列车灯光由远及近。中景越肩镜头,城市霓虹倒映在他们身后的玻璃上,暖橙色光与冷蓝色光交错,电影般的色彩分级。", "tags": [ "chinese", "cinematic", "emotional", "subway" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "jm-004", "model": "jimeng-3", "category": "social-viral", "title": "即梦抖音爆款:少女回眸", "prompt": "夏日午后,一位扎马尾的少女穿着白色 T 恤背对镜头走在金黄向日葵花田中,忽然回头朝镜头露出灿烂的笑容,发丝在风中轻舞。竖屏中景缓慢推进,逆光剪影效果,温暖电影感。", "tags": [ "chinese", "tiktok", "summer", "vertical" ], "params": { "aspect": "9:16", "duration": "6s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "jm-005", "model": "jimeng-3", "category": "fantasy-scifi", "title": "即梦仙侠:雪山修仙者", "prompt": "巍峨雪山之巅,一位身披白色狐裘的修仙者盘膝而坐,周身环绕淡蓝色灵气漩涡,长发随风飘扬。镜头从近景缓缓拉远,露出整座苍茫雪山,云海翻涌,远处可见悬浮的仙岛。仙侠 CG 风,水墨与写实融合。", "tags": [ "chinese", "xianxia", "fantasy", "epic" ], "params": { "aspect": "21:9", "duration": "10s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "jm-006", "model": "jimeng-3", "category": "comedy", "title": "即梦萌宠:橘猫偷食", "prompt": "一只圆滚滚的橘猫鬼鬼祟祟地跳上厨房料理台,伸出爪子去够一条放在盘子里的鱼,眼神警觉地左右张望,突然听到主人脚步声,立刻装出一副若无其事的样子坐下。室内自然光,可爱写实风格。", "tags": [ "chinese", "pet", "comedy", "tiktok" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "jm-007", "model": "jimeng-3", "category": "product-commercial", "title": "即梦电商:奶茶产品镜头", "prompt": "一杯透明杯装的桃子乌龙奶茶静静放在浅木桌上,桌面散落几朵新鲜桃花,背景是温暖的午后窗光透过纱帘,少女气息浓厚。镜头从俯视缓慢推进至侧面 45 度,杯壁水珠晶莹,奶茶层次清晰。商业广告级精致。", "tags": [ "chinese", "product", "food", "ecommerce" ], "params": { "aspect": "1:1", "duration": "6s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "jm-008", "model": "jimeng-3", "category": "nature", "title": "即梦自然:黄山日出云海", "prompt": "黄山之巅日出时分,一轮红日从云海中冉冉升起,金光洒在奇松怪石上,云海如波涛翻滚,远处山峦若隐若现。极广角航拍镜头,缓慢推进,国家地理纪录片质感,中式山水意境。", "tags": [ "chinese", "landscape", "sunrise", "aerial" ], "params": { "aspect": "21:9", "duration": "10s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "jm-009", "model": "jimeng-3", "category": "action", "title": "即梦武侠:竹林剑客", "prompt": "翠绿竹林中,一位白衣剑客手持长剑御风而立,剑尖凝结着细密水珠。突然一片竹叶飘落,剑客闪电般出剑将其精准切为两半。镜头从全景快速推进至剑尖特写,慢动作展现切割瞬间,武侠经典氛围。", "tags": [ "chinese", "wuxia", "bamboo", "slow-motion" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "jm-010", "model": "jimeng-3", "category": "quiet-moments", "title": "即梦生活:奶奶包饺子", "prompt": "厨房窗边,一位慈祥的奶奶正在认真包饺子,皱纹爬上眼角,手指娴熟地捏出花边,旁边小孙女好奇地踮脚观看。温暖柔和的下午阳光透过窗户,木桌上散落着面粉和饺子。亲切纪录片风格,中景平稳镜头。", "tags": [ "chinese", "family", "warm", "documentary" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "jm-011", "model": "jimeng-3", "category": "music-video", "title": "即梦音乐 MV:城市夜景跳舞", "prompt": "深圳市中心夜晚,霓虹闪烁的高楼天台上,一位青年街舞 dancer 在律动光影中展示流畅的 popping 动作,背景是无数闪烁的城市灯光和远处摩天大楼。低角度跟拍,霓虹紫粉蓝光影变化,运动模糊,潮流 MV 风格。", "tags": [ "chinese", "music-video", "dance", "neon" ], "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "jm-012", "model": "jimeng-3", "category": "creative", "title": "即梦创意:水墨变 3D 山水", "prompt": "一幅传统中国水墨山水画缓缓展开,画中的山峦、流水、孤舟逐渐脱离纸面,变成立体三维场景,最后镜头穿越进入画中世界,仿佛置身真实的山水之间。风格从水墨自然过渡到写实国风 CG,魔幻现实主义。", "tags": [ "chinese", "ink-wash", "transformation", "creative" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "即梦 AI", "url": "https://jimeng.jianying.com/", "tier": "official" } }, { "id": "kl-001", "model": "kling-3.0", "category": "product-commercial", "title": "香水广告(带角色配音 CTA)", "prompt": "Scene: Interior studio with smooth gradient lighting and a polished marble pedestal. Characters: A confident host (mid-30s, sleek black blazer) stands beside the bottle. Action: Close-up rotating perfume bottle, then the host introduces it enthusiastically, describing the scent notes with hand gestures. Extreme close-up of mist erupting in slow motion. Camera: Smooth orbit then push-in to extreme macro, ending on host's direct address. Audio & Style: Warm voiceover with subtle jazz bed, crisp atomizer hiss; luxury commercial polish.", "tags": [ "product", "perfume", "audio", "narrative", "5-layer" ], "params": { "aspect": "9:16", "duration": "10s" }, "source": { "name": "Atlabs AI", "url": "https://www.atlabs.ai/blog/kling-3-0-prompting-guide-master-ai-video-generation", "tier": "reputable-3rd-party" }, "notes": "示范 Kling 5 层公式:Scene → Characters → Action → Camera → Audio & Style" }, { "id": "kl-002", "model": "kling-3.0", "category": "product-commercial", "title": "香水广告(多镜头分镜 + 角色配音)", "prompt": "Scene: A sleek modern studio with soft gradient lighting and calm ambient music. Polished glass surface with subtle reflections. Characters: A lifestyle influencer in an elegant dress (early 30s, warm enthusiastic voice). Action: Shot 1 (0-4s) close-up on hands holding the perfume bottle, rotating it slowly to catch the light, glass gleaming. Shot 2 (4-8s) medium shot of the influencer, who says \"I have to show you my new signature scent because it is literally a dream.\" Shot 3 (8-11s) extreme close-up of perfume mist spraying in slow motion, particles catching light. Shot 4 (11-15s) return to influencer smiling at camera: \"Seriously, if you love vanilla and jasmine, you need this in your life. Link in my bio.\" Camera: Macro close-up to medium punch-in to extreme slow-mo, ending on direct address. Audio & Style: Soft atomizer spray SFX, elegant ambient music bed, luxury commercial polish.\nNegative: cartoonish, smooth plastic skin, floating limbs, text morphing, watermark.", "tags": [ "product", "perfume", "multi-shot", "audio", "dialog", "5-layer" ], "params": { "aspect": "9:16", "duration": "15s" }, "source": { "name": "Atlabs AI", "url": "https://www.atlabs.ai/blog/kling-3-0-prompting-guide-master-ai-video-generation", "tier": "reputable-3rd-party" }, "notes": "Atlabs 官方原文示例。Kling 3.0 多镜头分镜 + 单角色定向发声的标杆样式" }, { "id": "kl-003", "model": "kling-3.0", "category": "product-commercial", "title": "无线耳机磁悬旋转(图生视频)", "prompt": "Scene: Dark gradient background, cool white rim lighting, minimal floating stage. Characters: Matte black wireless headphones suspended at center. Action: A slow 180-degree orbit reveals cushion detail and brushed metal hinges; the headphones rotate gently as if weightless, subtle reflections shifting across the matte finish. Camera: Smooth horizontal orbit at constant speed, slight upward tilt at the midpoint, ending on a head-on hero frame. Audio & Style: Sub-bass swell synced to rotation, single bell tap on hero frame, premium product launch grade.\nNegative: visible rig, motion blur smearing logo, text morphing, watermark.", "tags": [ "product", "headphones", "image-to-video", "orbit", "minimal" ], "params": { "aspect": "1:1", "duration": "6s" }, "source": { "name": "Kling3 AI Guide", "url": "https://www.kling3-ai.com/blog/kling-3-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "改编自 kling3-ai.com Template 5。图生视频原则:只写运动方向 + 灯光细节,不重复描述产品已有外观" }, { "id": "kl-004", "model": "kling-3.0", "category": "product-commercial", "title": "蜂蜜淋松饼极致特写(图生视频)", "prompt": "Scene: Soft top-down kitchen window light, warm honey tones, shallow depth of field with creamy bokeh. Characters: A pair of pancakes stacked on a ceramic plate, a single hand holding a wooden honey dipper. Action: The dipper tilts; a thick golden ribbon of honey starts to fall, lands and pools across the top pancake, slowly cascading down the sides in viscous strands. Camera: Static extreme macro for the first beat, then a very slow push-in toward the pour point. Audio & Style: Thick liquid drip SFX, faint ambient kitchen tone, food-commercial texture and grain.\nNegative: artificial gloss, plastic-looking honey, sliding plate, watermark.", "tags": [ "food", "macro", "image-to-video", "asmr", "tactile" ], "params": { "aspect": "1:1", "duration": "5s" }, "source": { "name": "Kling3 AI Guide", "url": "https://www.kling3-ai.com/blog/kling-3-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "改编自 kling3-ai.com Template 4。极致 macro + 液体物理:Kling 强项,避免 dramatic lighting 这种笼统词" }, { "id": "kl-005", "model": "kling-3.0", "category": "dialogue-driven", "title": "厨房晨间混乱(四角色定向发声)", "prompt": "Scene: A busy kitchen at morning light, toast crumbs on the counter, coffee machine steaming. Cereal pouring SFX, kids running footsteps, backpack zippers. Characters: Mom (mid-30s, stressed, fast urgent voice), Little Daughter (6, crying), Older Brother (early teens, sarcastic), Dad (40s, sleepy amused). Action: Shot 1 (0-3s) wide shot, mom flipping toast — Mom: \"Shoes on! We're leaving in five minutes!\" Shot 2 (3-5s) hallway, daughter searching frantically — Little Daughter: \"I can't find my sweater!\" Shot 3 (5-7s) doorway, brother arms crossed — Older Brother: \"Because you never put it away.\" Shot 4 (7-9s) back to mom, heavy sigh — Mom (shouting): \"Nobody is fighting before 8 AM!\" Shot 5 (9-11s) dad walks in calmly with mug — Dad: \"Good morning, team.\" Shot 6 (11-13s) mom turns sharply, exhausted — Mom: \"Help.\" Camera: Tight handheld between rooms, single coverage per character, last beat held on mom. Audio & Style: Naturalistic kitchen ambience, no music, mumblecore family sitcom grade.\nNegative: synthetic dubbed voices, lip-sync drift, watermark.", "tags": [ "dialog", "family", "multi-character", "audio", "5-layer" ], "params": { "aspect": "16:9", "duration": "13s" }, "source": { "name": "Atlabs AI", "url": "https://www.atlabs.ai/blog/kling-3-0-prompting-guide-master-ai-video-generation", "tier": "reputable-3rd-party" }, "notes": "Atlabs 原文 Multi-Character Kitchen Drama 示例。Kling 3.0 杀手锏:4 个角色 + 角色定向发声 + 多镜头分镜,业界几乎独占" }, { "id": "kl-009", "model": "kling-3.0", "category": "cinematic", "title": "黄金时刻重逢(情感对话)", "prompt": "Scene: A park bench at golden hour, fall leaves drifting. Characters: Two reunited friends with a 10-year gap, both early 30s. Action: Alternating over-the-shoulder shots between both characters with nostalgic dialogue about missing each other, culminating in a moment of shared emotional recognition. Camera: Slow push-in on each speaker, ending on a held wide two-shot. Audio & Style: Soft piano underscore, ambient park sounds, warm cinematic color grade.", "tags": [ "reunion", "golden-hour", "audio", "dialog", "5-layer" ], "params": { "aspect": "16:9", "duration": "12s" }, "source": { "name": "Atlabs AI", "url": "https://www.atlabs.ai/blog/kling-3-0-prompting-guide-master-ai-video-generation", "tier": "reputable-3rd-party" } }, { "id": "kl-010", "model": "kling-3.0", "category": "cinematic", "title": "雨夜车内对话", "prompt": "Scene: Interior of a parked car at night with rain drumming on the roof. Characters: Driver (40s, weathered face) and passenger (20s, anxious). Action: Driver expresses hesitation while passenger avoids eye contact, exploration of misunderstanding through tense back-and-forth dialogue with rain intensifying throughout. Camera: Tight two-shots alternating with isolating singles, slight handheld drift. Audio & Style: Heavy rain ambience, no music, naturalistic dialogue.", "tags": [ "dialog", "rain", "tension", "audio", "5-layer" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Atlabs AI", "url": "https://www.atlabs.ai/blog/kling-3-0-prompting-guide-master-ai-video-generation", "tier": "reputable-3rd-party" } }, { "id": "kl-011", "model": "kling-3.0", "category": "cinematic", "title": "教练训斥鼓手(高强度训练)", "prompt": "Scene: A dimly lit jazz rehearsal room with cymbals catching key light. Characters: A drumming instructor (60s, intense) and a young drummer (20s, sweating). Action: Instructor critiques drummer's tempo with escalating fury through multiple close-ups on sweating hands, facial expressions, and cymbal crashes. Character stammers nervously while instructor demands precision through shouted corrections. Camera: Punchy quick cuts, snap zooms to drummer's face. Audio & Style: Loud cymbal hits, raw vocal performance, no music.", "tags": [ "tension", "audio", "drama", "5-layer" ], "params": { "aspect": "16:9", "duration": "12s" }, "source": { "name": "Atlabs AI", "url": "https://www.atlabs.ai/blog/kling-3-0-prompting-guide-master-ai-video-generation", "tier": "reputable-3rd-party" } }, { "id": "kl-012", "model": "kling-3.0", "category": "cinematic", "title": "电梯邂逅(克制情感)", "prompt": "Scene: An elevator at night, soft fluorescent ambient. Characters: Two strangers (late 20s) who discover shared music taste during an elevator encounter. Action: Subtle smiles and whispered recognition develop connection through minimal dialogue before doors open, suggesting possibility. Camera: Static two-shot with rack focus between faces. Audio & Style: Faint elevator hum, almost imperceptible music leak from one stranger's headphones.", "tags": [ "romance", "subtle", "minimal-dialog", "5-layer" ], "params": { "aspect": "9:16", "duration": "10s" }, "source": { "name": "Atlabs AI", "url": "https://www.atlabs.ai/blog/kling-3-0-prompting-guide-master-ai-video-generation", "tier": "reputable-3rd-party" } }, { "id": "kl-013", "model": "kling-3.0", "category": "image-to-video", "title": "棋手汗珠特写压力(图生视频)", "prompt": "Scene: Tournament chess hall, shallow depth of field, blurred wooden board and pieces in the background. Characters: A grandmaster mid-game (Magnus-Carlsen style, intense focus). Action: A bead of sweat slowly forms at the temple and traces down the side of his face; his eyes stay locked forward on the board, pupils micro-adjusting; jaw tightens, breath held, then released through the nose. Camera: Very slow push-in over the full duration, focus pinned to the eye and the sweat drop traveling across skin. Audio & Style: Near-silent room tone, distant chess-clock tick, a single low cello drone, suffocating concentration grade.\nNegative: cartoonish, smooth plastic skin, blinking out of rhythm, watermark.", "tags": [ "portrait", "image-to-video", "cinematic", "tension", "tactile" ], "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "Atlabs AI", "url": "https://www.atlabs.ai/blog/kling-3-0-prompting-guide-master-ai-video-generation", "tier": "reputable-3rd-party" }, "notes": "Atlabs 原文 Image-to-Video (Chess Example)。图生视频核心:只写动作(沿用图中已有的人物/环境)+ 显式 endpoint(汗珠到达脸颊)" }, { "id": "kl-018", "model": "kling-3.0", "category": "cinematic", "title": "审讯室对峙(双角色定向发声)", "prompt": "Scene: A sleek modern interrogation room with cold LED lighting, muted gray walls, a one-way glass window, security cameras blinking red. Low atmospheric suspense bed humming with deep bass drones. Characters: Lead Detective (40s, navy suit, controlled serious voice) and Prime Suspect (late 20s, sharp defensive voice). Action: The detective leans forward, hands resting calmly — Detective: \"Let's stop pretending.\" Suspect shifts in chair, tense — Suspect: \"I already told you everything.\" Detective slides a folder across the table with paper-scrape SFX — Detective (calm threatening): \"Then explain why your fingerprints are here.\" Suspect's breathing quickens — Suspect (trembling): \"That's impossible…\" Detective stands suddenly, chair scrapes back. Camera: Tight over-the-shoulder alternations, snap zoom on the folder, slow push-in on the suspect's eyes at the climax. Audio & Style: Rising pulse synth tightening into the final beat, room reverb on dialogue.\nNegative: smiling, soap-opera lighting, floating limbs, watermark.", "tags": [ "dialog", "tension", "audio", "thriller", "5-layer" ], "params": { "aspect": "16:9", "duration": "12s" }, "source": { "name": "fal.ai (Kling 3.0 Guide)", "url": "https://blog.fal.ai/kling-3-0-prompting-guide/", "tier": "official-platform" }, "notes": "fal.ai 官方 Kling 部署页原文示例 (Interrogation Room Scene)。双角色定向发声 + 物理音效(paper scrape / chair scrape)锚定真实感" }, { "id": "kl-019", "model": "kling-3.0", "category": "action", "title": "体育场冲场(队长鼓舞)", "prompt": "Scene: A stadium tunnel transitioning to a bright field. Characters: Players in team colors with the quarterback (captain) at the front. Action: Players depicted with backlit low angles gaining momentum, quarterback rallies team with commanding voice, then wide stadium shots with crowd and pyrotechnics. Camera: Tunnel POV → backlit low-angle hero shot → crane wide reveal. Audio & Style: Orchestral music building tension, crowd roar, captain's shouted voice line.", "tags": [ "sports", "stadium", "audio", "epic", "5-layer" ], "params": { "aspect": "16:9", "duration": "12s" }, "source": { "name": "Atlabs AI", "url": "https://www.atlabs.ai/blog/kling-3-0-prompting-guide-master-ai-video-generation", "tier": "reputable-3rd-party" } }, { "id": "kl-020", "model": "kling-3.0", "category": "action", "title": "都市跑酷夜行(FPV 无人机)", "prompt": "Scene: A narrow urban alley at night, magenta and cyan signage spilling onto wet asphalt, scaffolding overhead. Characters: A parkour runner in black sportswear, hood up. Action: The runner sprints into frame, vaults a row of dumpsters in one fluid motion, slides under low scaffolding without breaking stride, drags one gloved hand along a metal railing — sparks erupt and trail behind him. He launches off a stair edge and the camera barely keeps pace. Camera: Dynamic FPV drone shot, low and fast, chasing tightly behind the runner with quick yaw corrections through obstacles. Audio & Style: Breath rhythm in close mic, footfalls slapping wet ground, distant city hum, gritty handheld grade.\nNegative: slow motion (unless cued), sliding feet, floating limbs, watermark.", "tags": [ "action", "parkour", "fpv", "night", "audio" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Kling3 AI Guide", "url": "https://www.kling3-ai.com/blog/kling-3-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "改编自 kling3-ai.com Template 3。显式速度动词(sprints/vaults/slides)覆盖 Kling 默认慢动作倾向 — 09-kling-公式.md 第 6 节" }, { "id": "kl-035", "model": "kling-3.0", "category": "dialogue-driven", "title": "深夜厨房争吵(情感对白)", "prompt": "Scene: A dim kitchen late at night, only the refrigerator hum filling the silence, single overhead pendant casting a warm pool of light on a cluttered countertop. Characters: Exhausted Partner (30s, trembling frustrated voice) and Defensive Partner (30s, sharp defensive then regretful voice). Action: A plate is set down too hard with a sharp ceramic clink — Exhausted Partner: \"You never listen to me.\" The other partner turns immediately, eyes wide — Defensive Partner (shouting): \"Because you never stop blaming!\" Exhausted Partner exhales shakily — Exhausted Partner (voice cracking): \"I'm not blaming… I'm begging.\" Beat of silence. Defensive Partner sighs heavily, looks down — Defensive Partner (softly, regretful): \"I don't know how to fix this.\" Camera: Tight singles alternating between the two, brief two-shot on the final regret, soft handheld drift. Audio & Style: Refrigerator hum bed, ceramic clink hit, a single sad piano chord entering on the last beat.\nNegative: melodramatic underscoring, smiling, floating limbs, watermark.", "tags": [ "dialog", "domestic", "audio", "emotion", "5-layer" ], "params": { "aspect": "16:9", "duration": "12s" }, "source": { "name": "fal.ai (Kling 3.0 Guide)", "url": "https://blog.fal.ai/kling-3-0-prompting-guide/", "tier": "official-platform" }, "notes": "fal.ai 官方 Kling 部署页原文 (Kitchen Argument Scene)。展示 Kling 3.0 \"音画原生同步 + 情绪外化\" — 一个 piano chord 进入时机锁定情绪转折" }, { "id": "kl-036", "model": "kling-3.0", "category": "cinematic", "title": "霓虹拉面店推镜(35mm 胶片质感)", "prompt": "Scene: 一家狭长的霓虹拉面店,深夜,玻璃窗被蒸汽蒙上一层雾,外面街灯化作柔和的暖色 bokeh。 Characters: 一对年轻日本情侣并排坐在吧台前,碗里热气升腾。Action: 女子拿起筷子,慢慢卷起一缕面条,咬下时眼睛微微弯起;男子在旁边看着她,嘴角浮起一丝笑意;身后闪烁的洋红色霓虹招牌在雾窗上投下斑驳的反光。Camera: 缓慢的 dolly push-in,从店门口推进到吧台正前方,浅景深,焦点始终锁在女子的脸和那一缕面。Audio & Style: 碗筷轻碰声、安静的吸面声、远处城市嗡鸣,35mm 胶片暖颗粒,柔光高光晕染。\nNegative: 卡通化、塑料皮肤光泽、滑步、文字变形、水印。", "tags": [ "cinematic", "chinese", "intimate", "film-grain", "5-layer" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Kling3 AI Guide", "url": "https://www.kling3-ai.com/blog/kling-3-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "改编自 kling3-ai.com Ramen Shop 示例。展示 Kling 3.0 中文理解强项 — 全中文 5 层结构 + 胶片质感关键词。Kling 中文是业界并列第一" }, { "id": "kl-100", "model": "kling-3.0", "category": "cinematic", "title": "未来都市俯冲(5 要素结构基准)", "prompt": "Scene: A gleaming futuristic metropolis with towering glass skyscrapers and flying vehicles zooming between buildings. The city is bustling with activity as holographic advertisements illuminate the streets below. Neon lights reflect off wet pavement after a recent rainfall. Characters: Crowds of pedestrians moving in waves at the street level, distant figures in glass-walled offices. Action: The camera starts high above the city and gradually descends, moving through the layers of flying traffic to reveal the street level. Holographic ads shift across building faces. Camera: Slow vertical descent with parallax through traffic layers, ending in a wide street-level reveal. Audio & Style: Distant city hum, occasional flying-vehicle whoosh, rain remnants dripping; cinematic sci-fi color grade.\nNegative: blurry, distorted geometry, jerky camera, unnatural lighting.", "tags": [ "5-layer", "sci-fi", "cinematic", "establishing-shot", "city" ], "params": { "aspect": "16:9", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / EWBsU0fSgPs", "url": "https://www.youtube.com/watch?v=EWBsU0fSgPs", "tier": "reputable-3rd-party" }, "notes": "源自 EWBsU0fSgPs 5 要素结构示例 A1。重组为本仓库 5 层标签结构。" }, { "id": "kl-101", "model": "kling-3.0", "category": "image-to-video", "title": "山顶女性呼吸(图生视频)", "prompt": "Scene: Mountain peak vista with drifting clouds across an open sky; distant peaks fading into haze. Characters: A woman standing at the summit, hair and loose clothes catching the breeze, eyes closed in a long inhale. Action: The woman's hair and clothes gently move in the mountain breeze as she takes a deep breath of the fresh air. The clouds in the background drift slowly across the sky while distant birds soar between peaks. Camera: The camera gradually pulls back to reveal more of the majestic landscape surrounding her. Audio & Style: Crisp mountain wind, distant bird calls, soft ambient pad; awe-inspiring, golden-hour cinematic.", "tags": [ "5-layer", "image-to-video", "nature", "emotional", "pull-back" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / EWBsU0fSgPs", "url": "https://www.youtube.com/watch?v=EWBsU0fSgPs", "tier": "reputable-3rd-party" }, "notes": "源 A2。图生视频典型 — 主体在原图,只补 3 要素(动作+光线+摄像机)。" }, { "id": "kl-102", "model": "kling-3.0", "category": "image-to-video", "title": "咖啡杯静物追焦(图生视频)", "prompt": "Scene: A wooden window sill with soft window light spilling across the grain; outside, a busy street with people moving at varied paces. Characters: A single ceramic coffee cup with rising steam (the protagonist of the shot); blurred pedestrians beyond. Action: A coffee cup sits on a wooden window sill with soft light coming through the window. Outside, people walk by on the busy street, some rushing, others strolling. Camera: The camera slowly moves in a subtle arc around the cup, maintaining focus on the coffee cup while the background gradually blurs. Audio & Style: Muted urban hum, distant footsteps, subtle ambient cafe music; quiet still-life cinematic.", "tags": [ "5-layer", "image-to-video", "still-life", "arc-shot", "quiet" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / EWBsU0fSgPs", "url": "https://www.youtube.com/watch?v=EWBsU0fSgPs", "tier": "reputable-3rd-party" }, "notes": "源 A3。Subtle arc 镜头示范 — Kling 擅长的小幅运动。" }, { "id": "kl-103", "model": "kling-3.0", "category": "product-commercial", "title": "商务演示(Elements 静态环境)", "prompt": "Scene: A modern office space with floor-to-ceiling windows, neutral palette, soft fill lighting. Characters: A confident businessman in his 40s wearing a dark blazer, presenting energetically to an unseen audience. Action: The businessman confidently presents to an unseen audience in the modern office space, gesturing with his hands as he speaks. The office environment remains static while he moves naturally. Camera: The camera slowly zooms in on him as he delivers his presentation. Audio & Style: Subtle office ambience, clear voiced delivery, no music; clean corporate cinematic.", "tags": [ "5-layer", "elements", "corporate", "zoom-in", "presentation" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / EWBsU0fSgPs", "url": "https://www.youtube.com/watch?v=EWBsU0fSgPs", "tier": "reputable-3rd-party" }, "notes": "源 A4。Elements 模式典型 — 角色动,背景静。" }, { "id": "kl-104", "model": "kling-3.0", "category": "image-to-video", "title": "海滩黄金猎犬奔跑(Elements)", "prompt": "Scene: A bright sandy shoreline with rolling waves and seagulls drifting overhead; golden sun across wet sand. Characters: A golden retriever in mid-stride, ears bouncing, fur catching light. Action: The golden retriever runs playfully along the shoreline, kicking up sand and water as its paws hit the wet beach. Its fur and ears bounce with each stride as it chases after an unseen object. Waves gently roll onto the shore in the background while seagulls fly overhead. Camera: The camera pans smoothly following the dog from the side as it runs from left to right across the frame. Audio & Style: Waves, paws on wet sand, distant gull calls, gentle wind; warm summer cinematic.", "tags": [ "5-layer", "elements", "animal", "tracking-pan", "beach" ], "params": { "aspect": "16:9", "duration": "6s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / EWBsU0fSgPs", "url": "https://www.youtube.com/watch?v=EWBsU0fSgPs", "tier": "reputable-3rd-party" }, "notes": "源 A5。Side tracking pan,Kling 擅长的横向跟拍。" }, { "id": "kl-105", "model": "kling-3.0", "category": "action", "title": "Spider-Man vs Iron Man 极简对决", "prompt": "Scene: A barren wasteland under harsh open sky, dust and debris carried by wind. Characters: Spider-Man and Iron Man squaring off in combat stances. Action: Spider-Man and Iron Man fighting in a barren wasteland, epic and cinematic. Camera: Dynamic wide-to-medium tracking with brief whip transitions between hits. Audio & Style: Impact thuds, repulsor blasts, web shots; epic orchestral cinematic.", "tags": [ "5-layer", "action", "ip-style", "epic", "minimal-prompt" ], "params": { "aspect": "16:9", "duration": "6s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / 4_ZdSLA_OTQ", "url": "https://www.youtube.com/watch?v=4_ZdSLA_OTQ", "tier": "reputable-3rd-party" }, "notes": "源 B1。极简 prompt 也能出大片 — Kling 2.6 native audio 测试。" }, { "id": "kl-106", "model": "kling-3.0", "category": "dialogue-driven", "title": "咖啡馆对话(原生音频)", "prompt": "Scene: A cozy cafe with warm pendant lights, ambient chatter, ceramic clinks. Characters: Two friends in their 30s seated across a small wooden table, both with coffee cups. Action: Two friends sitting in a cozy cafe. One asks, \"Did you ever finish that novel?\" The other replies, \"Almost, I just can't seem to write the ending.\" Camera: Slow medium two-shot, soft punch-in on the speaker for each line. Audio & Style: Soft ambient cafe audio, warm lighting; understated indie-cinema tone.", "tags": [ "5-layer", "dialogue", "native-audio", "two-shot", "cafe" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / 4_ZdSLA_OTQ", "url": "https://www.youtube.com/watch?v=4_ZdSLA_OTQ", "tier": "reputable-3rd-party" }, "notes": "源 B2。Kling 2.6 原生音频对话 — 短对白,模型自动 lip-sync。" }, { "id": "kl-107", "model": "kling-3.0", "category": "cinematic", "title": "雨天咖啡馆女子独白", "prompt": "Scene: A cinematic rainy day cafe; rain splashes against the window with a cool blue-green tone; condensation on the glass; warm interior bulbs. Characters: A blonde woman, slightly damp from the rain, mid-30s, contemplative. Action: A cinematic rainy day cafe. Rain splashes against the window with a cool blue green tone. A blonde woman walks in slightly damp and sits down gazing directly at the camera and the woman says, \"Some moments don't disappear. They just wait for you to return.\" Camera: Medium frontal shot, very slow push-in as she sits and meets the lens. Audio & Style: Rain on glass, distant cafe murmur, low piano bed; melancholic cinematic.", "tags": [ "5-layer", "dialogue", "cinematic", "monologue", "rain" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / 4_ZdSLA_OTQ", "url": "https://www.youtube.com/watch?v=4_ZdSLA_OTQ", "tier": "reputable-3rd-party" }, "notes": "源 B3。直视镜头独白 + 情绪基调。原文 quote 完整保留。" }, { "id": "kl-108", "model": "kling-3.0", "category": "action", "title": "屋顶追逐落日跑酷", "prompt": "Scene: A skyline at sunset, narrow city rooftops with antennas and ducts, scattering pigeons. Characters: A hooded thief mid-sprint, lean and agile. Action: A rooftop chase at sunset. A thief sprints across narrow rooftops as a camera follows from behind. They leap between buildings as pigeons scatter in the sky. Camera: Fast dynamic camera movement following from behind, slight motion blur during leaps. Audio & Style: Audio of wind, footsteps, pigeon wings; fast cinematic atmosphere.", "tags": [ "5-layer", "action", "parkour", "follow-shot", "sunset" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / 4_ZdSLA_OTQ", "url": "https://www.youtube.com/watch?v=4_ZdSLA_OTQ", "tier": "reputable-3rd-party" }, "notes": "源 B4。Follow shot 跟拍典型 — 人物背对走开 = Kling 强项。" }, { "id": "kl-109", "model": "kling-3.0", "category": "cinematic", "title": "宇航员失重空间站", "prompt": "Scene: A spacious space station interior with metallic panels, floating instruments and slow caustic light through portholes. Characters: An astronaut in white suit, weightlessly drifting, helmet off, hair slightly floating. Action: Astronaut drifting weightlessly in a space station. Camera: Slow pan outwards. Audio & Style: Space station ambiance, distant hum, subtle systems beeps; serene sci-fi cinematic.", "tags": [ "5-layer", "sci-fi", "weightless", "slow-pan", "ambient" ], "params": { "aspect": "16:9", "duration": "6s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / 4_ZdSLA_OTQ", "url": "https://www.youtube.com/watch?v=4_ZdSLA_OTQ", "tier": "reputable-3rd-party" }, "notes": "源 B5。零对话情绪场 — Kling 3.0 不会随机加台词。" }, { "id": "kl-110", "model": "kling-3.0", "category": "action", "title": "Motion Control 击鼓(空气鼓)", "prompt": "Scene: A neutral studio backdrop with focused side lighting on a standing performer. Characters: A man in casual streetwear, mid-30s, energized expression, sweat sheen building. Action: The man wildly plays an invisible drum kit, striking the air with intense, energetic drumming motions. Camera: Locked medium-wide with subtle handheld breath; no cuts. Audio & Style: Faint cymbal whoosh, rhythmic body movement only; minimalist physical-comedy cinematic.", "tags": [ "5-layer", "motion-control", "action", "performance", "invisible-prop" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "MattVidPro / O-WFLK3em5I", "url": "https://www.youtube.com/watch?v=O-WFLK3em5I", "tier": "reputable-3rd-party" }, "notes": "源 C1。Motion Control 测试 — 动作迁移到静态角色图。" }, { "id": "kl-111", "model": "kling-3.0", "category": "action", "title": "Motion Control 功夫连击", "prompt": "Scene: An open martial-arts dojo with hardwood floor and soft top light, mild dust caught in beams. Characters: A man in plain training attire, focused jaw, breathing controlled. Action: The man performs a fast, continuous combo of kung fu fighting movements, striking the air with precise punches, sharp kicks, and flowing defensive blocks against invisible opponents. Camera: Locked wide-medium shot; one continuous take; no cuts. Audio & Style: Sharp gi snaps, exhales on impact, subtle floor scrape; classic martial-arts cinematic.", "tags": [ "5-layer", "motion-control", "kung-fu", "continuous-take", "dojo" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "MattVidPro / O-WFLK3em5I", "url": "https://www.youtube.com/watch?v=O-WFLK3em5I", "tier": "reputable-3rd-party" }, "notes": "源 C2。Motion Control 复杂连续动作。" }, { "id": "kl-112", "model": "kling-3.0", "category": "cinematic", "title": "东京霓虹雨夜独行男(摄像机聚焦)", "prompt": "Scene: A rain-soaked futuristic Tokyo street at night; neon signage reflected on the wet pavement; soft mist in the air. Characters: A lone man wearing a black trench coat, head slightly down, hands in pockets. Action: A lone man wearing a black trench coat walks through a rain-soaked futuristic Tokyo street at night. Neon reflections glowing across the wet pavement. Camera: Cinematic handheld camera slowly tracking toward him. Audio & Style: Rain hiss, distant traffic, faint signage hum; cyberpunk neo-noir.", "tags": [ "5-layer", "cinematic", "neo-noir", "tracking", "rain" ], "params": { "aspect": "16:9", "duration": "6s", "resolution": "1080p" }, "source": { "name": "Theoretically Media / cHpgSf7LKEE", "url": "https://www.youtube.com/watch?v=cHpgSf7LKEE", "tier": "reputable-3rd-party" }, "notes": "源 D2。Handheld 摄像机 + 聚焦人物 — Scene Context 技巧示范。" }, { "id": "kl-113", "model": "kling-3.0", "category": "action", "title": "拔枪眼神特写(Scene Context)", "prompt": "Scene: A dim interior with rim-lit silhouette and slight haze, mood between threat and inevitability. Characters: A man in a dark coat, mid-30s, sharp eyes, gun in waistband transitioning to hand. Action: He turns and raises a pistol; the lens commits to his eyes before the muzzle. Camera: camera zooms in on his eyes as he turns and raises a pistol. Audio & Style: Low rumble, single sharp metal click of the slide, breath holding; thriller cinematic.", "tags": [ "5-layer", "action", "scene-context", "zoom-in", "thriller" ], "params": { "aspect": "16:9", "duration": "4s", "resolution": "1080p" }, "source": { "name": "Theoretically Media / cHpgSf7LKEE", "url": "https://www.youtube.com/watch?v=cHpgSf7LKEE", "tier": "reputable-3rd-party" }, "notes": "源 D1。Scene Context 正确示范 — 加聚焦目标(on his eyes)。" }, { "id": "kl-114", "model": "kling-3.0", "category": "cinematic", "title": "商务女性咖啡馆(5 步框架)", "prompt": "Scene: A modern cafe with large windows and minimalist decor; warm morning sunlight streaming through; soft ambient cafe lighting. Characters: A confident businesswoman in her 30s wearing a cream blazer, holding a latte and a laptop. Action: A confident businesswoman in her 30s wearing a cream blazer sits by a window, slowly sipping her latte while checking emails on her laptop. Camera: Medium shot that slowly pulls back to reveal the full cafe setting. Audio & Style: Distant espresso machine hisses, light keyboard taps, soft jazz bed; cinematic color grading.", "tags": [ "5-layer", "5-step-framework", "cafe", "pull-back", "lifestyle" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "AI Conqueror / xoSbJFRnKMM", "url": "https://www.youtube.com/watch?v=xoSbJFRnKMM", "tier": "reputable-3rd-party" }, "notes": "源 E1。5 步框架标杆示例 — Subject + Action + Environment + Lighting + Camera。" }, { "id": "kl-115", "model": "kling-3.0", "category": "image-to-video", "title": "图生视频(5 步简化版,树下阅读)", "prompt": "Scene: A sun-dappled outdoor bench under a leafy tree (carried over from the input image). Characters: A reader holding an open book (carried over from the input image), peaceful expression. Action: Reading a book while occasionally turning pages and looking around. Camera: Slow push-in shot focusing on their peaceful expression. Audio & Style: Warm afternoon sunlight filtering through trees, distant bird call, paper rustle; soft pastoral cinematic.", "tags": [ "5-layer", "image-to-video", "5-step-framework", "push-in", "quiet" ], "params": { "aspect": "9:16", "duration": "5s", "resolution": "1080p" }, "source": { "name": "AI Conqueror / xoSbJFRnKMM", "url": "https://www.youtube.com/watch?v=xoSbJFRnKMM", "tier": "reputable-3rd-party" }, "notes": "源 E2。Image-to-Video 简化版 — 只补 3 项(动作+光线+摄像机)。" }, { "id": "kl-116", "model": "kling-3.0", "category": "product-commercial", "title": "LeBron 罚球眼部特写(Nano Banana Pro 广告)", "prompt": "Scene: An NBA arena interior at game tension; arena lights piercing across the player's face; sweat-shimmer on skin. Characters: LeBron James, locked in concentration during a free throw, eyes wide and steady. Action: An extreme close-up of LeBron James' eyes locked in focus during a free throw. The sweat is glistening on his skin. Arena lights sharply reflected on his pupils. Camera: Ultra shallow depth of field isolating the eyes with a creamy bokeh. Audio & Style: Crowd hush, single sneaker squeak, low arena rumble; high-end sport-ad cinematic.", "tags": [ "5-layer", "product-commercial", "sport", "extreme-close-up", "celebrity" ], "params": { "aspect": "16:9", "duration": "4s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / P7pH_1zFKbE", "url": "https://www.youtube.com/watch?v=P7pH_1zFKbE", "tier": "reputable-3rd-party" }, "notes": "源 F1。Nano Banana Pro + Kling 广告流程 — 眼部特写是核心情绪镜头。" }, { "id": "kl-117", "model": "kling-3.0", "category": "product-commercial", "title": "近静止运动镜头(汗珠细节)", "prompt": "Scene: Tight on the athlete's face; arena light reflections suspended on skin. Characters: The athlete, blinking once, gaze locked on the same far-off point. Action: Focused eyes, blinking once, looking at the same place. Sweat is like dripping down from one eye and we have some very subtle skin shimmer from the sweat. Camera: Keep the camera almost still with a slight natural handheld motion. Audio & Style: Soft breath, distant crowd, low sub bass; athletic-luxury commercial.", "tags": [ "5-layer", "product-commercial", "static-camera", "sweat-detail", "minimal-motion" ], "params": { "aspect": "16:9", "duration": "4s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / P7pH_1zFKbE", "url": "https://www.youtube.com/watch?v=P7pH_1zFKbE", "tier": "reputable-3rd-party" }, "notes": "源 F2。'almost still' camera 控制 — Kling 擅长的细微运动。" }, { "id": "kl-118", "model": "kling-3.0", "category": "product-commercial", "title": "球类过渡(Start/End Frame)", "prompt": "Scene: Mid-action court / pitch / center-court morphing into the next sport's location during the transition. Characters: The ball is the protagonist; player silhouettes blurred. Action: Ball transitions from basketball to football to tennis ball, occupying frame center across cuts. Camera: Camera constantly following the ball, keeping it centered. Do not let the ball go left or backwards. Keep the motion smooth and steady. Audio & Style: Whoosh, sport-specific impacts; energetic ad cinematic.", "tags": [ "5-layer", "product-commercial", "start-end-frame", "transition", "sport" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / P7pH_1zFKbE", "url": "https://www.youtube.com/watch?v=P7pH_1zFKbE", "tier": "reputable-3rd-party" }, "notes": "源 F3。球型过渡 Start/End Frame — 反向方向约束防止 AI 乱动。" }, { "id": "kl-119", "model": "kling-3.0", "category": "image-to-video", "title": "投篮出手动作(图生视频)", "prompt": "Scene: Free-throw line under arena lights (carried over from input image). Characters: The shooter (carried over from input), in shooting form. Action: Animate the player lifting the ball slightly higher and releasing it in a smooth controlled shooting motion. Camera: Static frame, very slight handheld breath. Audio & Style: Arena hush, sneaker squeak, net swish reserved for after release; broadcast cinematic.", "tags": [ "5-layer", "image-to-video", "sport", "subtle-motion", "release" ], "params": { "aspect": "16:9", "duration": "4s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / P7pH_1zFKbE", "url": "https://www.youtube.com/watch?v=P7pH_1zFKbE", "tier": "reputable-3rd-party" }, "notes": "源 F4。极简图生视频 prompt — 仅描述目标动作。" }, { "id": "kl-120", "model": "kling-3.0", "category": "cinematic", "title": "Kling O1 帧间转场(Constraint Sandwich)", "prompt": "Scene: A bridge between two reference frames (Start = frame A, End = frame B); ambient lighting matched. Characters: Whoever is in frame A and frame B; identity preserved. Action: Animate the transition between these frames. Keep the same character identity. Match the lighting. Cinematic slow drift. Camera: Slow drift along the implied motion vector between the two frames. Audio & Style: Subtle ambient pad; cinematic stillness.", "tags": [ "5-layer", "kling-o1", "start-end-frame", "transition", "preserve-identity" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "VWPrLusbFG4 (Artlist Kling O1)", "url": "https://www.youtube.com/watch?v=VWPrLusbFG4", "tier": "reputable-3rd-party" }, "notes": "源 G1。Kling O1 Start/End Frame 标准 prompt — 10s 1080p 支持。" }, { "id": "kl-121", "model": "kling-3.0", "category": "cinematic", "title": "霓虹小巷推镜(Kling 2.5 Turbo)", "prompt": "Scene: A neon-lit alleyway after a light rain; particles drifting in the colored shafts of light; mist near the floor. Characters: A female character in dark clothing, contemplative posture, faint reflective sheen on jacket. Action: Slow handheld push in on the female character. Particles drifting through the light. Camera: Slow handheld push-in, gentle organic sway. Audio & Style: Distant city low rumble, occasional dripping, subtle synth pad; dramatic atmosphere.", "tags": [ "5-layer", "kling-25-turbo", "neon", "handheld", "particles" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "VWPrLusbFG4 (Artlist Kling O1)", "url": "https://www.youtube.com/watch?v=VWPrLusbFG4", "tier": "reputable-3rd-party" }, "notes": "源 G2。Kling 2.5 Turbo 文生视频典型 — Drama via lighting + particles。" }, { "id": "kl-122", "model": "kling-3.0", "category": "cinematic", "title": "Kling O1 Next Shot(海鸥抢三明治)", "prompt": "Scene: A seaside picnic continuing from a reference video shot; same lighting and bench positions; gulls in midground. Characters: The woman holding a sandwich, the man mid-laugh; one seagull entering frame top-right. Action: Reference video source [video 1], generate the next shot. A seagull suddenly flies into the frame, steals the woman's sandwich. She is startled, and the man continues to laugh. Camera: Continuation of previous shot's framing; slight reactive whip toward the gull. Audio & Style: Wave wash, gull cry, man's laugh, woman's gasp; light comedy cinematic.", "tags": [ "5-layer", "kling-o1", "next-shot", "continuity", "comedy" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / m5WtZtI_MRg", "url": "https://www.youtube.com/watch?v=m5WtZtI_MRg", "tier": "reputable-3rd-party" }, "notes": "源 H1。Kling O1 Next Shot 模板 — 上传已有视频续写。" }, { "id": "kl-123", "model": "kling-3.0", "category": "viral-transform", "title": "地面裂开角色升起(Kling O1 特效)", "prompt": "Scene: An empty street or open field continuing from reference video 1; lighting matched; debris in the foreground. Characters: The character emerging from the hole in the ground, half body visible, dust on shoulders. Action: make the ground crack in video one. As dust and debris scatter, the character emerges from the hole in the ground as more dust comes out from within the hole. Camera: Low-angle hero shot fixed on the crack, slight rumble. Audio & Style: Low rumble, stone crack, debris clatter, sub bass impact; dramatic supernatural cinematic.", "tags": [ "5-layer", "kling-o1", "vfx", "transform", "low-angle" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / m5WtZtI_MRg", "url": "https://www.youtube.com/watch?v=m5WtZtI_MRg", "tier": "reputable-3rd-party" }, "notes": "源 H2。Kling O1 添加特效到已有视频 — 像 Photoshop generative fill。" }, { "id": "kl-124", "model": "kling-3.0", "category": "product-commercial", "title": "倒酒变产品镜头(Start/End Frame)", "prompt": "Scene: A bar countertop with neon signage in soft focus; reflections on the glass; a marble pedestal for the product hero. Characters: A hand pouring a drink (start frame), transitioning into a static product hero pose (end frame). Action: Pouring a drink into a glass transitions into a product shot. The neon lights glow and move in the background. Camera: Macro pour, motion follows the liquid, ending on a centered product hero composition. Audio & Style: Liquid pour, glass clink, soft neon hum, light bar music; modern premium ad cinematic.", "tags": [ "5-layer", "kling-o1", "start-end-frame", "product", "neon" ], "params": { "aspect": "9:16", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / m5WtZtI_MRg", "url": "https://www.youtube.com/watch?v=m5WtZtI_MRg", "tier": "reputable-3rd-party" }, "notes": "源 H3。Start/End Frame 产品广告流程 — 最常用功能。" }, { "id": "kl-125", "model": "kling-3.0", "category": "dialogue-driven", "title": "Multi-Shot 三镜头 Hollywood 自嘲", "prompt": "Scene: A neutral interior with cool key light, professional cinematic frame; same actor across all three angles. Characters: A man in casual sharp attire, mid-30s, deadpan delivery. Action: Shot 1: Man says \"Hollywood is officially cooked.\" Shot 2: Change angle to side profile. Man says \"because now you can have multi-shot videos.\" Shot 3: Change the angle again to a front closeup of the man. He says \"from a single image.\" Camera: Shot 1 front medium; Shot 2 side profile medium; Shot 3 front close-up. Audio & Style: Clear vocal recording, slight room tone, no music; tech-demo cinematic.", "tags": [ "5-layer", "multi-shot", "dialogue", "demo", "angle-cuts" ], "params": { "aspect": "16:9", "duration": "12s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / HTBfxEqDCdU", "url": "https://www.youtube.com/watch?v=HTBfxEqDCdU", "tier": "reputable-3rd-party" }, "notes": "源 I1。Kling 3.0 Multi-Shot 单图生成多角度的标杆 demo。" }, { "id": "kl-126", "model": "kling-3.0", "category": "viral-transform", "title": "OmniEdit 雪山换背景", "prompt": "Scene: Original character/product preserved from @video; environment swapped to a snowy mountain with falling flakes. Characters: Original subject of @video, identity and clothing preserved. Action: Restyle @video but change the background to be a snowy mountain. Camera: Preserve original camera framing and motion from @video. Audio & Style: Wind, snow crunch, distant mountain ambience; alpine cinematic.", "tags": [ "5-layer", "omniedit", "background-swap", "snow", "restyle" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / HTBfxEqDCdU", "url": "https://www.youtube.com/watch?v=HTBfxEqDCdU", "tier": "reputable-3rd-party" }, "notes": "源 I2。OmniEdit 背景替换最简模板。" }, { "id": "kl-127", "model": "kling-3.0", "category": "viral-transform", "title": "OmniEdit 风格图驱动重塑", "prompt": "Scene: Original scene preserved from @video; visual style globally driven by @image (color, light, grain, palette). Characters: Subjects of @video, posing preserved. Action: Restyle @video with the style of @image. Camera: Preserve original camera path. Audio & Style: Keep original ambient layer; style of @image dictates color grade, grain, and overall mood.", "tags": [ "5-layer", "omniedit", "style-transfer", "restyle", "image-reference" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / HTBfxEqDCdU", "url": "https://www.youtube.com/watch?v=HTBfxEqDCdU", "tier": "reputable-3rd-party" }, "notes": "源 I3。OmniEdit 风格图引导 — Kling 3.0 视频编辑核心能力。" }, { "id": "kl-128", "model": "kling-3.0", "category": "image-to-video", "title": "进门一刻 Start/End Frame", "prompt": "Scene: An apartment entryway transitioning from outside the closed door (start) to inside the lit foyer (end). Characters: A man in casual outerwear, mid-30s, slightly tired smile. Action: He grips the doorknob, turns it, and pushes through the door while saying \"I'm home.\" He steps inside and slides his hands from the doorknob to the edge of the door, looking into the room. Camera: Locked off frame matching both Start and End reference frames. Audio & Style: Doorknob click, hinge creak, distant kitchen sounds, warm voice; domestic cinematic.", "tags": [ "5-layer", "start-end-frame", "domestic", "dialogue", "transition" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Marlon Wallace / 0CjArtUh_Wg", "url": "https://www.youtube.com/watch?v=0CjArtUh_Wg", "tier": "reputable-3rd-party" }, "notes": "源 J1。Start/End Frame 经典开门 — 长动画视频流程。" }, { "id": "kl-129", "model": "kling-3.0", "category": "image-to-video", "title": "Scene Continuation(关门接镜)", "prompt": "Scene: Continuation of the previous shot's location; lighting and props preserved; transition to the kid mid-frame. Characters: A man (continuing from previous shot), and a child seated in the room. Action: A man closed the door. Camera turns slowly. Pans towards the kid. Camera: Slow pan from door to kid, continuation of previous shot's camera height. Audio & Style: Door click, soft footstep, child's faint hum; continuity-domestic cinematic.", "tags": [ "5-layer", "scene-continuation", "pan", "domestic", "two-character" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Marlon Wallace / 0CjArtUh_Wg", "url": "https://www.youtube.com/watch?v=0CjArtUh_Wg", "tier": "reputable-3rd-party" }, "notes": "源 J2。场景接续 — 上一镜最后帧 = 这一镜第一帧。" }, { "id": "kl-130", "model": "kling-3.0", "category": "cinematic", "title": "Multi-Shot 父子睡前三镜头", "prompt": "Scene: A child's bedroom with a soft nightlight, then transition to the hallway outside; warm low-key palette. Characters: A father in his 30s and his young son (4-6); both with subdued emotion. Action: Shot 1 (3s): A father tucks his young son into the bed. Gently brushes his hair. Both look sad. Shot 2 (4s): The father stands up, turns off the light, opens the door, and walks out into the hallway. The door slowly closes. Shot 3 (3s): @image2. The father leans against the wall besides the door and slowly slides down to sit on the floor. Camera: Shot 1 medium tracking; Shot 2 over-shoulder follow; Shot 3 wide static. Audio & Style: Soft breath, light switch click, door hush; emotional drama cinematic.", "tags": [ "5-layer", "multi-shot", "emotional", "scene-continuation", "father-son" ], "params": { "aspect": "16:9", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Marlon Wallace / 0CjArtUh_Wg", "url": "https://www.youtube.com/watch?v=0CjArtUh_Wg", "tier": "reputable-3rd-party" }, "notes": "源 J3。完整 3 镜头 Multi-Shot + @image2 引用 + 时长标注。" }, { "id": "kl-131", "model": "kling-3.0", "category": "action", "title": "F1 换胎 4 镜头(Multi-Shot)", "prompt": "Scene: A pit lane during a race; high-key arena lighting; brand-painted barriers; smoke from heated brakes. Characters: F1 car #front, full pit crew in matching team gear, three tire-change groups. Action: Shot 1: Grip shot on the front tire of the car as it pulls in — camera is mounted on the tire, spinning alongside it. Shot 2: The car pulls into the pit lane and stops. Wide shot. Shot 3: Three tire changes happen simultaneously — medium tracking shot. Shot 4: Grip shot on the front tire again as the car pulls away. Camera: Shot 1 tire-mounted POV; Shot 2 wide static; Shot 3 medium tracking; Shot 4 tire-mounted POV out. Audio & Style: Engine roar, impact wrench, crew shouts, tire screech; high-velocity sport cinematic.", "tags": [ "5-layer", "multi-shot", "action", "motorsport", "grip-shot" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / nUgx8ETiR1g", "url": "https://www.youtube.com/watch?v=nUgx8ETiR1g", "tier": "reputable-3rd-party" }, "notes": "源 K1。压力测试 — Multi-Shot 4 镜头 + 物理协调动作。" }, { "id": "kl-132", "model": "kling-3.0", "category": "action", "title": "PIT 追车雨夜电影感", "prompt": "Scene: A rain-slick highway at night; sirens rotating; spray rising off the asphalt. Characters: A police interceptor and the suspect's car; cop with focused stare. Action: Grip shot on the roof of the police car, right behind the sirens. Both cars make the turn. The cop car catches up and performs a PIT maneuver. Camera: Roof-mounted grip shot, slight rumble during impact. Audio & Style: Siren wail, rain hiss, tire screech, metal impact; raining; nighttime; cinematic.", "tags": [ "5-layer", "action", "police-chase", "grip-shot", "rain" ], "params": { "aspect": "16:9", "duration": "6s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / nUgx8ETiR1g", "url": "https://www.youtube.com/watch?v=nUgx8ETiR1g", "tier": "reputable-3rd-party" }, "notes": "源 K2。PIT 追车 — 单 prompt 高强度动作 + 天气 + 时段。" }, { "id": "kl-133", "model": "kling-3.0", "category": "image-to-video", "title": "Gen Z 影响者肖像生成", "prompt": "Scene: A clean studio setup, professional studio lighting, neutral backdrop with subtle gradient. Characters: An Asian woman in her 20s, Gen Z influencer styling, natural and beautiful, soft makeup. Action: A natural portrait shot of an Asian woman in her 20s. She looks like a 20-year-old Gen Z influencer, beautiful, natural looking with professional studio lighting. Camera: Tight portrait, very slight breath; could be locked off. Audio & Style: Studio quiet, soft pad if any; modern fashion-portrait cinematic.", "tags": [ "5-layer", "portrait", "image-to-video", "studio-lighting", "openart" ], "params": { "aspect": "9:16", "duration": "4s", "resolution": "1080p" }, "source": { "name": "OpenArt / h5kjDJrHw_g", "url": "https://www.youtube.com/watch?v=h5kjDJrHw_g", "tier": "reputable-3rd-party" }, "notes": "源 L1。OpenArt 角色生成 prompt — 用于建立角色库。" }, { "id": "kl-134", "model": "kling-3.0", "category": "image-to-video", "title": "多角色滑雪场合影(Elements 多引用)", "prompt": "Scene: A ski mountain summit (@image1) with crisp powder and bright sun; lift line in the distance. Characters: @Robin in ski outfit (@image2), @Leila in ski outfit (@image3); both posing. Action: @Robin and @Leila at the ski center, standing on top of the ski mountain at @image1. They are posing for a photo together. @Robin is wearing this ski outfit @image2, while @Leila is wearing this one @image3. Camera: Medium two-shot, slight push-in for warmth. Audio & Style: Wind on mountain, distant chair-lift hum, crunch of snow; bright winter cinematic.", "tags": [ "5-layer", "elements", "multi-character", "image-to-video", "winter" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "OpenArt / h5kjDJrHw_g", "url": "https://www.youtube.com/watch?v=h5kjDJrHw_g", "tier": "reputable-3rd-party" }, "notes": "源 L2。多角色 @ 引用 + 服装 @image 引用 + 环境 @image1。" }, { "id": "kl-135", "model": "kling-3.0", "category": "image-to-video", "title": "Robin 滑雪一致性视频", "prompt": "Scene: Ski mountain slope continuing from @image1 environment; powder snow, bright daylight. Characters: Robin, wearing the exact ski outfit from @image2; identity stable throughout. Action: Robin is going down the ski mountain on her skis, then she stops, and then she looks towards the camera. She stays perfectly consistent throughout the video. Wearing these exact clothes @image2. Camera: Tracking follow-shot then static medium on the stop and look. Audio & Style: Ski edge carve, wind, distant powder; sport-fashion cinematic.", "tags": [ "5-layer", "image-to-video", "consistency", "sport", "tracking" ], "params": { "aspect": "16:9", "duration": "10s", "resolution": "1080p" }, "source": { "name": "OpenArt / h5kjDJrHw_g", "url": "https://www.youtube.com/watch?v=h5kjDJrHw_g", "tier": "reputable-3rd-party" }, "notes": "源 L3。Kling 2.6 Start/End Frame 一致性 — 锚定服装。" }, { "id": "kl-136", "model": "kling-3.0", "category": "dialogue-driven", "title": "Brooklyn Bridge 情侣争吵(文生视频)", "prompt": "Scene: Brooklyn Bridge at night; warm sodium and cool moonlight; foot traffic faint in the background. Characters: A couple in their late 20s; the woman near tears, the man clenching jaw. Action: \"Don't do this. I am begging you.\" Shot 1: Brooklyn Bridge, couple arguing, close-up on woman's face, 5s. Shot 2: reverse angle on man, 5s. Shot 3: wide establishing shot, 5s. Camera: Shot 1 close-up; Shot 2 reverse close-up; Shot 3 wide establishing. Audio & Style: Ambient bridge traffic, footsteps, no music, no background dialogue.", "tags": [ "5-layer", "multi-shot", "dialogue", "text-to-video", "drama" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Theoretically Media / tqZ0JuUevwA", "url": "https://www.youtube.com/watch?v=tqZ0JuUevwA", "tier": "reputable-3rd-party" }, "notes": "源 M1。文生视频 + 3 镜头分镜 + 显式 no music no background dialogue 约束。" }, { "id": "kl-137", "model": "kling-3.0", "category": "cinematic", "title": "时间戳结构(7 种结构 #2)", "prompt": "Scene: An outdoor exterior with a key character holding a recording device; sky overhead. Characters: A solo character in casual clothing, neutral mood transitioning to wonder. Action: At 0-3s the character is approached as the camera zooms in. At 3-5s gaze drops to the recording device in his hands. At 5-8s gaze lifts to the sky. Camera: [0-3s] the camera zooms in towards him. [3-5s] the camera tilts down to show him looking at this recording device. [5-8s] the camera tilts back up again to see him looking up towards the sky. Audio & Style: Wind, distant ambient, light electronic pad; contemplative cinematic.", "tags": [ "5-layer", "timestamp", "camera-tilt", "contemplative", "structure-2" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / zzBmvzR-URg", "url": "https://www.youtube.com/watch?v=zzBmvzR-URg", "tier": "reputable-3rd-party" }, "notes": "源 N1。时间戳 prompt 结构 — 单场景多动作编排。" }, { "id": "kl-138", "model": "kling-3.0", "category": "cinematic", "title": "切镜结构(7 种结构 #3)", "prompt": "Scene: An alien terrain with a parked spaceship in the background; dust on the boots; cinematic film stock look. Characters: An astronaut in a battered suit. Action: The astronaut turns and walks towards a spaceship. Cut to the boots of the astronaut walking along the terrain. Camera: Wide medium for the turn; cut to extreme close-up of the boots; both rooted in same film aesthetic. Audio & Style: Footstep crunch, suit-mic breath, distant low rumble; cinematic film stock.", "tags": [ "5-layer", "cut-to", "sci-fi", "structure-3", "boots" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / zzBmvzR-URg", "url": "https://www.youtube.com/watch?v=zzBmvzR-URg", "tier": "reputable-3rd-party" }, "notes": "源 N2。Cut to 切镜结构 — 风格须一致避免跳变。" }, { "id": "kl-139", "model": "kling-3.0", "category": "action", "title": "无声扣扳机(负向 prompt 示范)", "prompt": "Scene: A tense intimate interior; deliberate silence broken only by breath; key light low. Characters: A character at a moment of restraint, hand on a weapon but choosing not to fire. Action: The character's hand tightens, then relaxes; deliberate stillness. Camera: Slow push-in to the hand, then tilt up to the face. Audio & Style: completely silent and quiet and no gunshots, no clicking, no trigger; suspense cinematic.", "tags": [ "5-layer", "negative-prompt", "silence", "suspense", "structure-7" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / zzBmvzR-URg", "url": "https://www.youtube.com/watch?v=zzBmvzR-URg", "tier": "reputable-3rd-party" }, "notes": "源 N3。负向 prompt #7 — 直接告诉 AI 不想要什么比描述正确更高效。" }, { "id": "kl-140", "model": "kling-3.0", "category": "image-to-video", "title": "Husky 跑上膝头(动作描述胜过图像描述)", "prompt": "Scene: Living room interior preserved from input image; sofa, cushions, soft window light. Characters: The woman seated on the sofa (from input image); a husky puppy entering frame. Action: A husky puppy runs onto the woman's lap. Camera: Static frame; tiny natural breath only. Audio & Style: Paws on floor, puppy whine, soft laugh; warm domestic cinematic.", "tags": [ "5-layer", "image-to-video", "trigger-entrance", "animal", "static-camera" ], "params": { "aspect": "9:16", "duration": "4s", "resolution": "1080p" }, "source": { "name": "AI Search / vqwaJZ5RVxU", "url": "https://www.youtube.com/watch?v=vqwaJZ5RVxU", "tier": "reputable-3rd-party" }, "notes": "源 O1。Action description > image description — Kling 图生视频核心原则。" }, { "id": "kl-141", "model": "kling-3.0", "category": "image-to-video", "title": "跪地祈祷(多动作并发)", "prompt": "Scene: An interior preserved from input image; soft directional light. Characters: A figure (from input image), respectful expression. Action: kneeling and praying. Camera: Static medium; possibly micro push-in. Audio & Style: Soft breath, quiet ambient room tone; reverent cinematic.", "tags": [ "5-layer", "image-to-video", "multi-action", "minimal-prompt", "religious" ], "params": { "aspect": "9:16", "duration": "4s", "resolution": "1080p" }, "source": { "name": "AI Search / vqwaJZ5RVxU", "url": "https://www.youtube.com/watch?v=vqwaJZ5RVxU", "tier": "reputable-3rd-party" }, "notes": "源 O2。多动作 prompt 同时执行 — 标准模式 + 最大 Relevance。" }, { "id": "kl-142", "model": "kling-3.0", "category": "cinematic", "title": "多摄像机序列机械臂(3.0 评测)", "prompt": "Scene: A studio with a robotic-arm camera rig; subject locked to a precise framing target; controlled environment. Characters: A subject (face, body) holding pose for the rig. Action: Camera performs a fast lateral pass left to right. Then a brief crash push in to the face. Then a quick pull back and the fast pass back right to the left. Camera: Robotic arm. Very snappy acceleration. Audio & Style: Servo motor whir, sharp whoosh, deep low-end punch on the crash; technical demo cinematic.", "tags": [ "5-layer", "camera-sequence", "robotic-arm", "demo", "kling-3.0" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / jQvHqKrGhy8", "url": "https://www.youtube.com/watch?v=jQvHqKrGhy8", "tier": "reputable-3rd-party" }, "notes": "源 P1。Kling 3.0 多步摄像机序列无形变测试 — rolling shutter 效果保留。" }, { "id": "kl-143", "model": "kling-3.0", "category": "image-to-video", "title": "咬洋葱圈 Zoom + 慢动作", "prompt": "Scene: A casual eatery interior preserved from input image; warm light on the food. Characters: A girl reaching for a single fried onion ring; lips emphasized. Action: A girl grabs one fried onion. The handheld camera zooms in and stops on the lips just as she takes the bite. Add a brief slow motion to the bite itself. Camera: Handheld zoom-in stopping on lips, brief slow-mo on the bite. Audio & Style: Crunch, breath, ambient diner; tasty food-cinema cinematic.", "tags": [ "5-layer", "image-to-video", "zoom-in", "slow-motion", "food" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / jQvHqKrGhy8", "url": "https://www.youtube.com/watch?v=jQvHqKrGhy8", "tier": "reputable-3rd-party" }, "notes": "源 P2。Less-is-More + 关键动作 brief slow-mo — Kling 3.0 自主细节补全。" }, { "id": "kl-144", "model": "kling-3.0", "category": "dialogue-driven", "title": "三段 Multi-Shot 围栏旁夜对话", "prompt": "Scene: An apartment courtyard at night; warm window light spilling onto a wooden fence; cool sky behind. Characters: A young man and woman, both early 30s, emotional tension. Action: Scene 1: Medium shot of the young man and woman standing by the fence at night. The man is speaking earnestly, his face illuminated by the warm light from the apartment building. Scene 2: Reverse angle over the shoulder shot, focusing on the woman. She listens with a serious expression, then looks down and sighs before looking back up to respond. Scene 3: Over the shoulder shot focusing on the man, he listens as the woman responds. Camera: Scene 1 medium; Scene 2 OTS woman; Scene 3 OTS man. Audio & Style: Crickets, distant TV through a window, soft voices; emotional drama cinematic.", "tags": [ "5-layer", "multi-shot", "dialogue", "ots", "drama" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / jQvHqKrGhy8", "url": "https://www.youtube.com/watch?v=jQvHqKrGhy8", "tier": "reputable-3rd-party" }, "notes": "源 P3。Three-shot dialogue with consistent face — Kling 3.0 跨场景一致性 + lip sync 测试。" }, { "id": "kl-145", "model": "kling-3.0", "category": "cinematic", "title": "霓虹直升机疾速穿城", "prompt": "Scene: A neon-lit city at night with glowing exhaust fumes and motion blur on the streets below; deep blue and fiery orange color grade. Characters: A futuristic helicopter, lights blazing, in fast forward motion. Action: Create a high resolution video depicting cinematic shots of a futuristic helicopter racing through a neon lit city at night. The scene is color graded with deep blues and fiery oranges, emphasizing the speed and intensity. Camera: The camera follows from a low angle rear perspective with fast motion blur, vibrant exhaust fumes that trail behind, amplifying the dynamic energy of the scene. Audio & Style: Rotor thwap, sub-bass rumble, neon hum; cyberpunk cinematic.", "tags": [ "5-layer", "cinematic", "vehicle", "follow-shot", "neon" ], "params": { "aspect": "16:9", "duration": "6s", "resolution": "1080p" }, "source": { "name": "Kevin Stratvert / anACVZDNw6o", "url": "https://www.youtube.com/watch?v=anACVZDNw6o", "tier": "reputable-3rd-party" }, "notes": "源 Q1。完整 5 要素结构 + low angle rear perspective。" }, { "id": "kl-146", "model": "kling-3.0", "category": "image-to-video", "title": "卡通小鸟图生视频", "prompt": "Scene: A bright, colorful kid-friendly scene preserved from input image. Characters: Two cartoon birds, vibrant feathers, expressive eyes. Action: Cartoon style, two birds are chirping and flapping their wings. Camera: Locked off, micro reactive. Audio & Style: Cute chirps, light cartoon sting; colorful kids vibe.", "tags": [ "5-layer", "image-to-video", "cartoon", "kids", "minimal-prompt" ], "params": { "aspect": "16:9", "duration": "4s", "resolution": "1080p" }, "source": { "name": "Kevin Stratvert / anACVZDNw6o", "url": "https://www.youtube.com/watch?v=anACVZDNw6o", "tier": "reputable-3rd-party" }, "notes": "源 Q2。极简风格描述 + 'Colorful kids vibe' 关键词。" }, { "id": "kl-147", "model": "kling-3.0", "category": "image-to-video", "title": "雾林模特肖像启动帧", "prompt": "Scene: A foggy forest with a faint distant castle silhouette visible through the mist; cinematic lighting; shallow depth of field. Characters: A model in moody portrait pose, natural skin texture, soft-focus eyes. Action: The model breathes slowly, fog slightly shifts behind her; very subtle movement. Camera: Locked close-up; shallow depth of field; minimal breath in the frame. Audio & Style: Forest wind, distant crows, subtle pad; moody photorealistic.", "tags": [ "5-layer", "image-to-video", "portrait", "fog", "subtle-motion" ], "params": { "aspect": "9:16", "duration": "5s", "resolution": "1080p" }, "source": { "name": "AI Master / _kYAkogj9WE", "url": "https://www.youtube.com/watch?v=_kYAkogj9WE", "tier": "reputable-3rd-party" }, "notes": "源 R1。Source 用作主图描述 + 衍生为图生视频。" }, { "id": "kl-148", "model": "kling-3.0", "category": "cinematic", "title": "维多利亚海雾码头(GPT 优化)", "prompt": "Scene: A fog-shrouded pier at dusk; weathered wood slick with moisture; gas lanterns casting warm amber pools of light; a large sailing vessel anchored in the distance. Characters: A woman alone in Victorian-era attire — a dark green coat with brass buttons, a fitted corset bodice, and a long layered skirt with subtle plaid. Her hair is swept up loosely with a few strands escaping in the sea breeze. Action: A woman stands alone on a fog-shrouded pier at dusk. She wears Victorian-era attire — a dark green coat with brass buttons, a fitted corset bodice, and a long layered skirt with subtle plaid. Her hair is swept up loosely with a few strands escaping in the sea breeze. She gazes out toward a large sailing vessel anchored in the distance, her expression contemplative. Camera: slow push-in, medium close shot, focus shifts to her eyes at the end. Audio & Style: Sea wind, distant gull, lapping water, faint creak of rigging; cinematic color grade, desaturated blues and greens with warm lantern highlights.", "tags": [ "5-layer", "period-piece", "victorian", "gpt-optimized", "push-in" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / cGTBzed4S4w", "url": "https://www.youtube.com/watch?v=cGTBzed4S4w", "tier": "reputable-3rd-party" }, "notes": "源 S2。GPT 训练官方文档输出 — 详尽 5 要素 + focus shift 镜头。" }, { "id": "kl-149", "model": "kling-3.0", "category": "image-to-video", "title": "孙悟空愤怒表情(极简动作)", "prompt": "Scene: A studio backdrop preserved from MidJourney Sun Wukong source frame; detailed fur and accessories; muted cinematic colors; Cinestill film grain. Characters: The Monkey King Sun Wukong, intense expression. Action: Animate an angry expression. Camera: Static camera. Audio & Style: Subtle low growl, faint wind, distant temple bell; Cinestill film grain, muted cinematic.", "tags": [ "5-layer", "image-to-video", "static-camera", "expression", "minimal-prompt" ], "params": { "aspect": "16:9", "duration": "4s", "resolution": "1080p" }, "source": { "name": "Adventures with Ai / Mrq34YqIlV0", "url": "https://www.youtube.com/watch?v=Mrq34YqIlV0", "tier": "reputable-3rd-party" }, "notes": "源 T2。Kling Professional 模式 + 'subtle motion / static camera / slowly' 关键词。" }, { "id": "kl-150", "model": "kling-3.0", "category": "product-commercial", "title": "Atomic 香水黄金时刻旁白", "prompt": "Scene: A golden-hour Parisian rooftop or balcony; soft warm rim light on glass and fabric. Characters: Atomic bottle as hero product; an elegant hand entering frame in the second beat. Action: Voiceover delivers the line as the bottle is approached and lifted slowly. \"Bathe in the golden hour. Atomic, crafted with the quiet power of Paris. Unleash the moment one sip at a time.\" Camera: Macro hero on bottle, slow push-in then tilt to hand. Audio & Style: Soft narration, light Parisian piano motif, ambient golden-hour wind; luxury commercial cinematic.", "tags": [ "5-layer", "product-commercial", "narration", "golden-hour", "perfume" ], "params": { "aspect": "9:16", "duration": "10s", "resolution": "1080p" }, "source": { "name": "z84WQAn6U0I (Kling 50 prompts)", "url": "https://www.youtube.com/watch?v=z84WQAn6U0I", "tier": "reputable-3rd-party" }, "notes": "源 U1。产品广告 + 人声旁白 — Kling 3.0 多语言对话支持。" }, { "id": "kl-151", "model": "kling-3.0", "category": "viral-transform", "title": "废土战场 AI 元素合成(Kling O1 VFX)", "prompt": "Scene: An open field; while the man walks, an apocalyptic battlefield with scattered junk should be composited around him as shown in image 1; smoke and debris. Characters: The man from Video 1, identity and walk cycle preserved. Action: Video 1: man walks through a field. An apocalyptic battlefield with scattered junk around the field as shown in image 1. Camera: Preserve Video 1 framing; subtle parallax of background junk. Audio & Style: Wind, distant rumble, debris scrape, faint distant explosions; post-apocalyptic cinematic.", "tags": [ "5-layer", "kling-o1", "vfx-compositing", "post-apocalyptic", "video-to-video" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / eMDDDdoUgT4", "url": "https://www.youtube.com/watch?v=eMDDDdoUgT4", "tier": "reputable-3rd-party" }, "notes": "源 V1。AE RotoBrush 3 + Nano Banana Pro + Kling O1 合成。" }, { "id": "kl-152", "model": "kling-3.0", "category": "viral-transform", "title": "刺客双肩动物中世纪(Motion Library 多元素)", "prompt": "Scene: A Third Crusade, late 1100s, Middle East medieval setting; dusty palette, sandstone backdrop. Characters: An assassin in dark robes; an orange cat on his left shoulder; a cute animal on his right shoulder. Action: Close above the assassin with the orange cat sitting on his left shoulder, and the cute animal sitting on his right shoulder. Camera: A slow, consistent pan inward. Audio & Style: Wind, distant horse hooves, faint chanting; historic-fantasy cinematic.", "tags": [ "5-layer", "elements-library", "multi-element", "historical", "pan-in" ], "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Curious Refuge / yUBGI_kUm-g", "url": "https://www.youtube.com/watch?v=yUBGI_kUm-g", "tier": "reputable-3rd-party" }, "notes": "源 W1。Elements 库多元素组合 — 同一 prompt 调 3 个以上元素。" }, { "id": "kl-153", "model": "kling-3.0", "category": "image-to-video", "title": "图生视频官方摄像机跟随(Spells)", "prompt": "Scene: A street or path environment preserved from input image; lighting and time-of-day from input image. Characters: The woman (carried over from input image), walking with steady pace. Action: the woman walks while the camera is following her. Camera: Camera follows her continuously, kept behind or alongside; per Kling official Spells structure. Audio & Style: Footsteps on the path surface, ambient environment audio matching the input image setting.", "tags": [ "5-layer", "image-to-video", "official-structure", "follow-shot", "minimal-prompt" ], "params": { "aspect": "9:16", "duration": "5s", "resolution": "1080p" }, "source": { "name": "i-Y9I33e_WA (Ultimate Cinematic Prompting Guide)", "url": "https://www.youtube.com/watch?v=i-Y9I33e_WA", "tier": "reputable-3rd-party" }, "notes": "源 X1。Kling 官方 Spells 结构 — 图生视频:主体 + 动作 + 摄像机在末尾。" }, { "id": "kl-154", "model": "kling-3.0", "category": "cinematic", "title": "悬崖女探险家(文生视频官方示例)", "prompt": "Scene: A vast canyon with red rock formations; warm golden hour sunlight casting long shadows; morning light, adventure atmosphere. Characters: A female adventurer in a brown leather jacket and worn canvas pants, weathered confidence in her stance. Action: A female adventurer in a brown leather jacket and worn canvas pants stands at the edge of a cliff overlooking a vast canyon. She raises her right hand to shade her eyes and scans the horizon. Camera: Aerial view. Audio & Style: Mountain wind, distant raptor call, low ambient; cinematic.", "tags": [ "5-layer", "text-to-video", "official-example", "aerial", "adventure" ], "params": { "aspect": "16:9", "duration": "6s", "resolution": "1080p" }, "source": { "name": "i-Y9I33e_WA (Ultimate Cinematic Prompting Guide)", "url": "https://www.youtube.com/watch?v=i-Y9I33e_WA", "tier": "reputable-3rd-party" }, "notes": "源 X2。Kling 文生视频官方示例 — shot type 'Aerial view' 写在 prompt 最前。" }, { "id": "kl-155", "model": "kling-3.0", "category": "viral-transform", "title": "Samurai 3×3 多角度网格(角色一致性源)", "prompt": "Scene: A quiet rural road near rice fields, with distant mountains and wooden villages fading into mist; evening wind moving fabric. Characters: A lone samurai wearing worn indigo robes and partial armor, a sheathed katana resting at his side. Action: A cinematic 3×3 grid presenting multiple camera angles. Each frame shows a distinct shot: extreme close-up, medium shot, over-the-shoulder, wide shot, high-angle top-down, low-angle, profile, three-quarter rear view, and full back view. Camera: Nine static framings as specified; consistent across all panels. Audio & Style: Natural lighting, muted earth tones, soft fog, minimal color saturation, shallow depth of field, cinematic framing. Consistent subject appearance across all frames, film-still aesthetic, professional cinematography reference sheet style, ultra-realistic photography. Images stacked together with no grid lines and no borders. No Text.", "tags": [ "5-layer", "3x3-grid", "character-reference", "samurai", "consistency" ], "params": { "aspect": "1:1", "duration": "image-grid", "resolution": "4K" }, "source": { "name": "OA5g63cBp7A (Seamless AI Films from One Prompt)", "url": "https://www.youtube.com/watch?v=OA5g63cBp7A", "tier": "reputable-3rd-party" }, "notes": "源:网格生成 prompt (Nano Banana / Kling 配合) — 用于建立 Elements 源图。" }, { "id": "kl-156", "model": "kling-3.0", "title": "AI 网红品牌合作视频 — Starbucks 推近镜头(Dan Kieft Pack 原文)", "prompt": "The camera pushes in slowly from a medium to a close-up as she lifts the Starbucks cup to take a sip, then lowers it with a soft smile while glancing toward the side. Her hand stays in her pocket, posture relaxed but confident. The street behind her softly blurs as warm lights inside the cafe begin to glow subtly.", "tags": [ "brand-collab", "influencer", "push-in", "character-consistency", "image-to-video", "starbucks-style" ], "category": "product-commercial", "params": { "aspect": "9:16", "duration": "5s", "resolution": "1080p" }, "source": { "name": "YouTube fxiDS2C9XYA · Dan Kieft AI Influencer Prompt Pack (237K views)", "url": "https://www.youtube.com/watch?v=fxiDS2C9XYA", "tier": "reputable-3rd-party", "note": "Google Docs Prompt Pack 原文(角色卡 + 图像 + 视频 prompt 串联),对应 methodology/20 §3.6" } }, { "id": "lt-001", "model": "ltx-video", "category": "nature", "title": "LTX 单段 prose 样板:岩岸浪花", "prompt": "The turquoise waves crash against the dark, jagged rocks of the shore, sending white foam spraying into the air. The scene is dominated by the stark contrast between the bright blue water and the dark, almost black rocks. The water is a clear, turquoise color, and the waves are capped with white foam. The rocks are dark and jagged, and they are covered in patches of green moss. The shore is lined with lush green vegetation, including trees and bushes. In the background, there are rolling hills covered in dense forest. The sky is cloudy, and the light is dim.", "tags": [ "nature", "single-paragraph", "ltx-official", "real-time" ], "params": { "aspect": "16:9", "duration": "10s", "resolution": "1216x704" }, "source": { "name": "LTX-Video Official", "url": "https://huggingface.co/Lightricks/LTX-Video", "tier": "official-platform" }, "notes": "官方推荐:单段 4-8 句自然流动 prose,present tense,literal precise" }, { "id": "lt-002", "model": "ltx-video", "title": "LTX screenplay 格式:小镇新闻报道", "prompt": "EXT. SMALL TOWN STREET – MORNING – LIVE NEWS BROADCAST. The shot opens on a news reporter standing in front of cordoned-off cars, yellow caution tape fluttering behind him. He wears a navy blue jacket, holds a black microphone in his right hand, and addresses the camera with a serious expression. He has short brown hair, light skin, and visible stubble. The camera holds a medium shot, focused on his face and upper body. Warm early sun reflects off the chrome of a vintage station wagon parked behind him; a uniformed officer walks past in the background and pauses near the tape. The reporter glances briefly off-screen to his right, then back to the camera, continuing to speak. The scene is captured in real-life footage, with natural ambient sound from the street.", "tags": [ "screenplay", "dialogue", "ltx-official", "narrative" ], "category": "dialogue-driven", "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Lightricks LTX-Video", "url": "https://huggingface.co/Lightricks/LTX-Video-0.9.7-dev", "tier": "official-platform" } }, { "id": "lt-003", "model": "ltx-video", "category": "quiet-moments", "title": "LTX 日常生活:居家办公男青年", "prompt": "Young man in blue t-shirt sitting at desk with laptop in home office setting, typing on keyboard with natural hand movements then leaning back in chair to think, realistic everyday video style with soft natural window lighting from the left, medium shot from front at chest height, plant in background slightly out of focus, calm productive afternoon atmosphere.", "tags": [ "everyday", "lifestyle", "ltx-official", "natural-light" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "LTX-Video", "url": "https://huggingface.co/Lightricks/LTX-Video", "tier": "official-platform" }, "notes": "LTX 擅长写实日常场景,不擅长奇幻夸张" }, { "id": "lt-004", "model": "ltx-video", "category": "image-to-video", "title": "LTX I2V 只描述运动", "prompt": "She turns to camera and smiles gently, wind catches her hair lifting it from her shoulders. Warm afternoon light brightens her face. The camera holds steady, then slowly drifts forward an inch.", "tags": [ "image-to-video", "motion-only", "ltx-official" ], "params": { "aspect": "9:16", "duration": "5s" }, "source": { "name": "LTX-Video", "url": "https://huggingface.co/Lightricks/LTX-Video", "tier": "official-platform" }, "notes": "I2V 铁律:只描述运动 + 光线变化,不重描述图中已有内容" }, { "id": "lt-005", "model": "ltx-video", "category": "action", "title": "LTX 街头跑酷一气呵成", "prompt": "A young man in a grey hoodie sprints down a narrow alleyway between brick buildings, vaults a chain-link fence in one fluid motion, lands on his feet, and continues running without breaking stride. The handheld camera tracks alongside him at the same pace, the late afternoon sun casting long shadows across the asphalt, dust kicked up by his footfalls catching the warm light.", "tags": [ "action", "parkour", "tracking", "natural" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "LTX-Video Community", "url": "https://huggingface.co/Lightricks/LTX-Video", "tier": "official-platform" } }, { "id": "lt-006", "model": "ltx-video", "title": "LTX 实时雨中咖啡", "prompt": "A woman in her thirties sits alone at a window seat in a cafe, watching rain streak down the glass. She has light skin, long brown hair tied loosely, and wears a cream-colored sweater over a white collared shirt. The camera remains stationary, framing her in a medium close-up from the side. She wraps both hands around a steaming ceramic cup, lifts it slowly to her lips, and takes a small sip; her eyes stay fixed on the rain outside. A small drop of condensation slides down the cup. Outside the window, blurred figures with umbrellas pass on the wet sidewalk, their reflections catching the warm interior light of the cafe. The lighting is soft and natural, predominantly from the gray daylight diffused through the rain on the glass. The scene appears to be real-life footage.", "tags": [ "slice-of-life", "rain", "cafe", "natural" ], "category": "cinematic", "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Lightricks LTX-Video", "url": "https://huggingface.co/Lightricks/LTX-Video-0.9.7-dev", "tier": "official-platform" } }, { "id": "lt-007", "model": "ltx-video", "category": "nature", "title": "LTX 森林晨光", "prompt": "Morning sunlight breaks through the canopy of a dense pine forest, illuminating drifting mist between the trees. A small stream runs across a bed of mossy stones in the foreground. Birds flutter between branches. The camera moves slowly forward at chest height, dew on leaves catches and refracts the golden light, the entire scene breathes with quiet morning calm.", "tags": [ "nature", "forest", "morning", "atmospheric" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "LTX-Video", "url": "https://comfyanonymous.github.io/ComfyUI_examples/ltxv/", "tier": "official" } }, { "id": "lt-008", "model": "ltx-video", "category": "open-source", "title": "LTX 本地部署测试 prompt", "prompt": "A close-up of weathered hands shaping wet clay on a spinning potter's wheel. Water drips from the artist's fingers, the clay slowly rises into a tall narrow vessel under steady pressure. Soft window light from the left, sawdust particles drift in the beam. The wheel hums quietly, hands work with practiced calm.", "tags": [ "craft", "test", "open-source", "macro" ], "params": { "aspect": "1:1", "duration": "8s" }, "source": { "name": "LTX-Video", "url": "https://github.com/Lightricks/LTX-Video", "tier": "official-platform" }, "notes": "适合本地部署首次测试:写实材质 + 自然光 + 缓慢动作,LTX 最稳" }, { "id": "mo-001", "model": "mochi-1", "category": "physics-sim", "title": "Mochi 玻璃慢镜破碎", "prompt": "A slow-motion shot of a wine glass shattering on a polished concrete floor, fragments of glass spinning outward through the air, droplets of red wine arcing in slow trails, captured from a low angle with shallow depth of field.", "tags": [ "physics", "slow-motion", "destruction", "mochi-official" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Genmo", "url": "https://www.genmo.ai/blog/mochi-1-a-new-sota-in-open-text-to-video", "tier": "official" }, "notes": "Mochi 强项:5.4s 慢动作 + 物理细节准确" }, { "id": "mo-002", "model": "mochi-1", "category": "craft", "title": "Mochi 街头粉笔画延时", "prompt": "A time-lapse of a street artist creating a large chalk mural on a city sidewalk, the colorful image of a phoenix gradually emerging stroke by stroke, pedestrians passing by in fast motion, late afternoon sunlight shifting across the pavement.", "tags": [ "timelapse", "art", "street", "mochi-official" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Genmo", "url": "https://www.genmo.ai/blog/mochi-1-a-new-sota-in-open-text-to-video", "tier": "official" } }, { "id": "mo-003", "model": "mochi-1", "category": "portrait", "title": "Mochi 阳光林间骑行", "prompt": "A young woman with long brown hair rides a vintage cream-colored bicycle down a dappled forest path, motion-blur style, sunlight filtering through the leaves above creating moving patterns of light on her face and the path ahead, joyful relaxed expression, late summer afternoon.", "tags": [ "lifestyle", "motion-blur", "forest", "mochi-official" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Genmo Replicate", "url": "https://replicate.com/genmoai/mochi-1/examples", "tier": "official-platform" } }, { "id": "mo-004", "model": "mochi-1", "category": "nature", "title": "Mochi 黄金时刻瀑布", "prompt": "A majestic waterfall cascading down mossy rocks in a lush rainforest, golden hour sunlight filtering through the canopy, mist rising from the base of the falls catching the warm light in rainbow refractions, slow cinematic pan from left to right revealing the full height of the cascade.", "tags": [ "nature", "waterfall", "cinematic", "mochi-official" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Genmo", "url": "https://replicate.com/genmoai/mochi-1/examples", "tier": "official-platform" } }, { "id": "mo-005", "model": "mochi-1", "category": "action", "title": "Mochi 草原猎豹冲刺", "prompt": "A cheetah sprinting across an open savanna at full speed, muscles flexing with each stride, dust kicking up behind it in plumes, dramatic wide shot, 4K wildlife documentary style, golden grass parting under its paws, the African horizon stretching endlessly behind.", "tags": [ "wildlife", "action", "documentary", "mochi-official" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Genmo", "url": "https://replicate.com/genmoai/mochi-1/examples", "tier": "official-platform" } }, { "id": "mo-006", "model": "mochi-1", "category": "open-source", "title": "Mochi 本地部署测试 prompt", "prompt": "A close-up of fresh coffee beans being poured slowly from a wooden scoop into a glass jar, beans bouncing and settling with realistic weight, soft natural light from a kitchen window, shallow depth of field, warm domestic atmosphere.", "tags": [ "product", "macro", "test", "open-source" ], "params": { "aspect": "1:1", "duration": "5s" }, "source": { "name": "Genmo", "url": "https://github.com/genmoai/mochi", "tier": "official-platform" } }, { "id": "mo-007", "model": "mochi-1", "category": "quiet-moments", "title": "Mochi 海边孤独漫步", "prompt": "A figure in a long grey coat walks alone along an empty beach at sunset, footprints trailing in the wet sand behind them, the ocean to their left reflecting orange and pink sky, wind whipping the coat tails sideways, distant lighthouse silhouette on the horizon, melancholic atmospheric mood.", "tags": [ "atmospheric", "beach", "sunset", "melancholic" ], "params": { "aspect": "21:9", "duration": "5s" }, "source": { "name": "Genmo Community", "url": "https://www.genmo.ai/", "tier": "official" } }, { "id": "mo-008", "model": "mochi-1", "category": "fantasy-scifi", "title": "Mochi 太空舱观察", "prompt": "An astronaut floating inside a spacecraft cabin, looking out a large circular window at Earth below, blue and white cloud patterns slowly rotating, soft cabin lighting reflecting on her visor, her gloved hand pressed against the glass, awe in her expression, weightless hair drifting around her face.", "tags": [ "sci-fi", "astronaut", "earth", "atmospheric" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Genmo", "url": "https://www.genmo.ai/", "tier": "official" } }, { "id": "pk-001", "model": "pika-2.5", "title": "Pikaffects: Melt 巧克力融化", "prompt": "A chocolate cat figurine on white marble slowly collapses inward, melt effect, glossy chocolate flows outward in viscous streams, soft warm key light from upper left, close-up macro shot, smooth 24fps slow motion, no morphing of other objects in frame.", "tags": [ "pikaffects", "melt", "viral", "macro", "asmr" ], "category": "pikaffects", "params": { "aspect": "1:1", "duration": "5s", "resolution": "1080p" }, "source": { "name": "fal.ai · Pika Effects (v1.5) docs", "url": "https://fal.ai/models/fal-ai/pika/v1.5/pikaffects", "tier": "official-platform" } }, { "id": "pk-002", "model": "pika-2.5", "title": "Pikaffects: Explode 苹果爆裂", "prompt": "A glossy red apple resting on dark wood explodes outward in slow motion, explode effect, particles and juice spray symmetrically toward camera, black background, dramatic side rim lighting, high-speed photography aesthetic, single subject in focus, no morphing.", "tags": [ "pikaffects", "explode", "satisfying", "macro", "slow-motion" ], "category": "pikaffects", "params": { "aspect": "1:1", "duration": "4s", "resolution": "1080p" }, "source": { "name": "fal.ai · Pika Effects (v1.5) docs", "url": "https://fal.ai/models/fal-ai/pika/v1.5/pikaffects", "tier": "official-platform" } }, { "id": "pk-003", "model": "pika-2.5", "title": "Pikaffects: Squish 蛋糕被压", "prompt": "A pink frosted layer cake on a pastel plate, squish effect, a giant translucent hand descends and gently flattens it, soft deformation with sprinkles bouncing outward, even key light, top-down 9:16 framing, single subject focus, no morphing of background.", "tags": [ "pikaffects", "squish", "asmr", "viral", "pastel" ], "category": "pikaffects", "params": { "aspect": "9:16", "duration": "5s", "resolution": "1080p" }, "source": { "name": "fal.ai · Pika Effects (v1.5) docs", "url": "https://fal.ai/models/fal-ai/pika/v1.5/pikaffects", "tier": "official-platform" } }, { "id": "pk-004", "model": "pika-2.5", "title": "Pikaffects: Inflate 气球膨胀", "prompt": "A small rubber duck on a wooden table, inflate effect, the duck swells to four times its size while staying upright, latex surface taut and glossy, soft daylight from window left, locked off camera, single subject focus, no morphing of table or background.", "tags": [ "pikaffects", "inflate", "viral", "playful" ], "category": "pikaffects", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "fal.ai · Pika Effects (v1.5) docs", "url": "https://fal.ai/models/fal-ai/pika/v1.5/pikaffects", "tier": "official-platform" } }, { "id": "pk-005", "model": "pika-2.5", "title": "Pikaffects: Levitate 物体悬浮", "prompt": "A vintage hardcover book resting on an oak desk, levitate effect, the book slowly rises six inches and rotates 30 degrees while pages flutter open, warm tungsten desk lamp from right, mid shot at eye level, single subject focus, no morphing of other desk items.", "tags": [ "pikaffects", "levitate", "mystical", "books" ], "category": "pikaffects", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "fal.ai · Pika Effects (v1.5) docs", "url": "https://fal.ai/models/fal-ai/pika/v1.5/pikaffects", "tier": "official-platform" } }, { "id": "pk-006", "model": "pika-2.5", "title": "Pikaffects: Cake-ify 物品变蛋糕", "prompt": "A leather handbag on a marble pedestal, cake-ify effect, the leather transforms layer by layer into pink fondant cake with white piping and rosettes, soft pastel studio light, slow rotating shot, single subject focus, no morphing of pedestal or background.", "tags": [ "pikaffects", "cake-ify", "viral", "transformation" ], "category": "pikaffects", "params": { "aspect": "1:1", "duration": "5s", "resolution": "1080p" }, "source": { "name": "fal.ai · Pika Effects (v1.5) docs", "url": "https://fal.ai/models/fal-ai/pika/v1.5/pikaffects", "tier": "official-platform" } }, { "id": "pk-007", "model": "pika-2.5", "title": "Pika 电影感模板(雨夜霓虹)", "prompt": "A lone figure in a dark trench coat walks down a rain-soaked Tokyo alley at night, neon signs reflect on wet asphalt in pink and cyan, steady tracking shot from behind at medium-wide framing, 35mm cinematic look with shallow depth of field, soft handheld micro-shake, single subject focus.", "tags": [ "cinematic", "neon", "rain", "tracking-shot" ], "category": "cinematic", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Imagine.art · Pika 2.2 features", "url": "https://www.imagine.art/features/pika-2-2", "tier": "reputable-3rd-party" } }, { "id": "pk-008", "model": "pika-2.5", "title": "Pika 中国风万花筒(Imagine.art 官方示例)", "prompt": "Panning into a chinese style kaleidoscope scene made of glass, the girl turns around and looks at the camera, ornate symmetrical patterns shift gently behind her, soft studio backlight rim, single subject focus, no morphing of kaleidoscope elements.", "tags": [ "chinese-style", "portrait", "kaleidoscope", "cinematic" ], "category": "cinematic", "params": { "aspect": "9:16", "duration": "5s", "resolution": "1080p" }, "source": { "name": "Imagine.art · Pika 2.2 features (官方示例 prompt)", "url": "https://www.imagine.art/features/pika-2-2", "tier": "reputable-3rd-party" } }, { "id": "pk-009", "model": "pika-2.5", "title": "Pika 产品镜头模板(香水瓶)", "prompt": "A clear glass perfume bottle on a polished black marble surface, slow 360° orbit around the bottle, mist emerges from the atomizer in a thin curl, soft studio key light from upper left with subtle specular highlights on glass edges, single subject focus, commercial photography aesthetic, no morphing.", "tags": [ "product", "perfume", "orbit", "studio", "commercial" ], "category": "product-commercial", "params": { "aspect": "1:1", "duration": "5s", "resolution": "1080p" }, "source": { "name": "fal.ai · Pika v2.5 text-to-video", "url": "https://fal.ai/models/fal-ai/pika/v2.2/text-to-video", "tier": "official-platform" } }, { "id": "pk-010", "model": "pika-2.5", "title": "Pika 角色面部表情(短片爆款)", "prompt": "Close-up of a young woman in her late 20s laughing softly, her shoulders shake gently, eyes crinkle at the corners, golden hour backlight catches stray hairs, shallow depth of field with face in sharp focus, locked off camera, character animation focus, no morphing of facial features.", "tags": [ "character", "portrait", "facial-animation", "viral", "close-up" ], "category": "portrait", "params": { "aspect": "9:16", "duration": "5s", "resolution": "1080p" }, "source": { "name": "fal.ai blog · Pika models on fal (character animation 2.5 强项)", "url": "https://blog.fal.ai/pika-models-now-available-on-fal/", "tier": "official-platform" } }, { "id": "pk-011", "model": "pika-2.5", "title": "Pika POV 第一视角咖啡店", "prompt": "First person POV walking into a cozy morning coffee shop, hands push open wooden door at frame edges, golden window light bathes wood tables, baristas working at espresso machines visible mid-distance, gentle handheld stabilized motion forward, single immersive shot, no morphing.", "tags": [ "pov", "social-viral", "coffee-shop", "morning", "handheld" ], "category": "social-viral", "params": { "aspect": "9:16", "duration": "5s", "resolution": "1080p" }, "source": { "name": "fal.ai · Pika v2.5 text-to-video", "url": "https://fal.ai/models/fal-ai/pika/v2.2/text-to-video", "tier": "official-platform" } }, { "id": "pk-012", "model": "pika-2.5", "title": "Pikaframes 关键帧过渡(春樱到秋叶)", "prompt": "Keyframe 1: a single cherry blossom branch in full pink bloom against soft blue spring sky. Keyframe 2: the same branch in autumn with deep amber leaves against grey overcast sky. Smooth Pikaframes transition between the two keyframes, locked composition, the branch stays centered, scene morphs naturally between seasons.", "tags": [ "pikaframes", "keyframe", "nature", "seasonal-transition", "transformation" ], "category": "nature", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "fal.ai · Pika v2.2 Pikaframes (Keyframe Image-to-Video) API docs", "url": "https://fal.ai/models/fal-ai/pika/v2.2/pikaframes/api", "tier": "official-platform" } }, { "id": "rw-001", "model": "runway-gen4", "category": "cinematic", "title": "银衣女子玻璃走廊(Runway 官方样板)", "prompt": "A woman in a silver coat walks through a long glass corridor at night. Reflections drift across the walls as she turns slightly toward the camera. Medium tracking shot, cool blue light, cinematic realism.", "tags": [ "cinematic", "tracking", "noir", "runway-official" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Runway", "url": "https://help.runwayml.com/hc/en-us/articles/39789879462419-Gen-4-Video-Prompting-Guide", "tier": "official" }, "notes": "Gen-4 官方推荐结构:subject + action + setting + camera + motion + style" }, { "id": "rw-002", "model": "runway-gen4", "category": "video-edit", "title": "Aleph 编辑:换更广镜头 + 冷调重光", "prompt": "Change the camera to a wider angle that reveals the full body. Re-light the scene with colder blue tones and stronger reflections on the glass. Keep the subject's movement natural and preserve the original pacing.", "tags": [ "aleph", "edit", "re-light", "re-frame" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Runway", "url": "https://help.runwayml.com/hc/en-us/articles/43277392678803-Aleph-Prompting-Guide", "tier": "official" }, "notes": "Aleph 编辑思维:说什么要改 + 什么要稳定 + 转换的程度" }, { "id": "rw-003", "model": "runway-gen4", "category": "video-edit", "title": "Aleph 编辑:变黄金时刻光照", "prompt": "Make it golden hour. Warm sunlight from camera left creating long shadows. Preserve the subject and composition exactly.", "tags": [ "aleph", "edit", "re-light", "golden-hour" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Runway", "url": "https://help.runwayml.com/hc/en-us/articles/43277392678803-Aleph-Prompting-Guide", "tier": "official" } }, { "id": "rw-004", "model": "runway-gen4", "category": "video-edit", "title": "Aleph 编辑:加街灯到夜晚", "prompt": "Add streetlights to make this look like night. Cast warm orange glow on the wet pavement. Keep everything else unchanged.", "tags": [ "aleph", "edit", "add", "night" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Runway", "url": "https://help.runwayml.com/hc/en-us/articles/43277392678803-Aleph-Prompting-Guide", "tier": "official" } }, { "id": "rw-005", "model": "runway-gen4", "category": "video-edit", "title": "Aleph 编辑:移除背景路人", "prompt": "Remove all background pedestrians while keeping the main subject and architecture intact. Maintain the original camera movement.", "tags": [ "aleph", "edit", "remove" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Runway", "url": "https://help.runwayml.com/hc/en-us/articles/43277392678803-Aleph-Prompting-Guide", "tier": "official" } }, { "id": "rw-006", "model": "runway-gen4", "category": "video-edit", "title": "Aleph 编辑:换风格成水彩动画", "prompt": "Re-style this clip as a hand-painted watercolor animation. Soft brush textures, flowing edges, warm palette. Preserve all motion and timing exactly.", "tags": [ "aleph", "edit", "re-style", "watercolor" ], "params": { "aspect": "16:9", "duration": "5s" }, "source": { "name": "Runway", "url": "https://help.runwayml.com/hc/en-us/articles/43277392678803-Aleph-Prompting-Guide", "tier": "official" } }, { "id": "rw-007", "model": "runway-gen4", "category": "cinematic", "title": "雨夜霓虹独行者", "prompt": "A lone figure in a dark trench coat walks slowly down a rain-soaked alley at night, neon signs reflecting on wet pavement. Slow tracking shot from behind. High contrast, deep shadows, cyberpunk noir.", "tags": [ "noir", "neon", "rain", "tracking" ], "params": { "aspect": "21:9", "duration": "8s" }, "source": { "name": "Runway", "url": "https://help.runwayml.com/hc/en-us/articles/39789879462419-Gen-4-Video-Prompting-Guide", "tier": "official" } }, { "id": "rw-008", "model": "runway-gen4", "category": "portrait", "title": "工作室肖像柔光特写", "prompt": "Close-up portrait of a woman with curly red hair, soft natural window light from the left, slight smile, eyes meeting camera, shallow depth of field, cinematic skin tones, subtle hair movement.", "tags": [ "portrait", "soft-light", "close-up" ], "params": { "aspect": "1:1", "duration": "6s" }, "source": { "name": "Runway", "url": "https://help.runwayml.com/hc/en-us/articles/39789879462419-Gen-4-Video-Prompting-Guide", "tier": "official" } }, { "id": "rw-009", "model": "runway-gen4", "title": "瀑布旁的摩托追逐", "prompt": "A motorcycle speeds along a winding mountain road beside a tall waterfall. Low tracking shot follows from the side at high speed. Spray and mist from the falls catch the late afternoon light. Dynamic motion blur on the foreground rocks, cinematic realism, golden hour color palette. The road curves out of frame to the right as the rider leans into the turn. Natural handheld camera shake. Observational documentary feel.", "tags": [ "action", "motorcycle", "waterfall", "tracking" ], "category": "action", "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Runway Gen-4 Video Prompting Guide", "url": "https://help.runwayml.com/hc/en-us/articles/39789879462419-Gen-4-Video-Prompting-Guide", "tier": "official" } }, { "id": "rw-010", "model": "runway-gen4", "title": "冰岛极简风景", "prompt": "A vast Icelandic black sand beach at dusk. A single basalt column stands in the foreground, rough textured stone catching the last directional light. Gentle waves wash the shore in repeating cycles. Static wide shot, locked-off composition. Muted blue-grey color palette, hyperrealistic textures, slow ambient drift in the distant water. Atmospheric stillness, no morphing, observational stillness.", "tags": [ "landscape", "iceland", "minimal", "atmospheric" ], "category": "nature", "params": { "aspect": "21:9", "duration": "8s" }, "source": { "name": "Runway Gen-4 Video Prompting Guide", "url": "https://help.runwayml.com/hc/en-us/articles/39789879462419-Gen-4-Video-Prompting-Guide", "tier": "official" } }, { "id": "rw-011", "model": "runway-gen4", "title": "手表光影漂浮", "prompt": "A luxury silver watch suspended against a deep black background, slowly rotating in place. A crane shot moves smoothly downwards along the watch, revealing the bezel, dial, and crown in sequence. Single key light from upper right creates crisp specular highlights along the polished steel. Macro detail focus on the dial. Subtle particles drift in the surrounding darkness. Cinematic stillness, no morphing, no extra elements.", "tags": [ "product", "watch", "luxury", "macro" ], "category": "product-commercial", "params": { "aspect": "1:1", "duration": "8s" }, "source": { "name": "Runway Gen-4 Video Prompting Guide", "url": "https://help.runwayml.com/hc/en-us/articles/39789879462419-Gen-4-Video-Prompting-Guide", "tier": "official" } }, { "id": "rw-012", "model": "runway-gen4", "category": "fantasy-scifi", "title": "未来宇航员触摸全息", "prompt": "An astronaut in a sleek white spacesuit reaches out to touch a glowing holographic star map floating in front of her. Inside a dimly lit spaceship cockpit. Medium close-up, slow push-in. Blue and amber light reflections on her visor. Realistic textures, sci-fi cinematic.", "tags": [ "sci-fi", "astronaut", "hologram" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Runway", "url": "https://help.runwayml.com/hc/en-us/articles/39789879462419-Gen-4-Video-Prompting-Guide", "tier": "official" } }, { "id": "sd-001", "model": "seedance-2.0", "title": "⭐ 经典 · 纯爱叙事:少年少女书桌目光相遇(15s timecode)", "prompt": "15秒唯美纯爱叙事镜头,电影级写实画质,2.35:1画幅。0-4秒:暖黄色调,午后阳光透过百叶窗洒在木质书桌上,形成细微的丁达尔效应,少女趴在桌上浅睡,微风吹动她耳边的碎发和书页,镜头极缓推向侧脸特写;5-10秒:镜头切换至对面坐着的少年,他正出神地看着少女,眼神清澈溢满温柔,发现少女睫毛微动,立刻慌乱地低头假装看书,指尖局促地揉搓书角;11-15秒:大特写两人目光撞在一起的瞬间,少女羞涩抿嘴,少年露出腼腆笑容,背景虚化成梦幻的光斑,音效是轻微的蝉鸣和翻书声。禁止出现任何水印、字幕、文字。", "tags": [ "signature-classic", "narrative", "romance", "youth", "cinematic", "timecode", "chinese", "tyndall-effect", "shallow-focus" ], "category": "cinematic", "params": { "aspect": "2.35:1", "duration": "15s", "resolution": "1080p" }, "source": { "name": "@LufzzLiz (岚叔) · X tweet — 仓库主作者经典作品 (201K 浏览 · 671 赞 · 778 bookmarks)", "url": "https://x.com/LufzzLiz/status/2029817696448303419", "video_url": "https://video.twimg.com/amplify_video/2029816303431962624/vid/avc1/1472x632/YAxaxw5HfX7TGmeQ.mp4?tag=21", "tier": "reputable-3rd-party", "note": "本仓库经典样板 — 中文 timecode prompt + 丁达尔效应 + 双角色情绪外化 + 浅景深背景虚化的完整范例,推荐作为入门必读" } }, { "id": "sd-002", "model": "seedance-2.0", "category": "product-commercial", "title": "食品教程特写(Dreamina 官方 Viral 模板 #3)", "prompt": "Short recipe tutorial with appetizing macro shots. Subject: a pair of clean hands and the key ingredient (cream pouring into hot espresso). Action: the cream column descends in slow motion, forming a vortex on the dark coffee surface; the hands gently lift the cup and rotate it toward the lens. Scene: a styled cafe counter with warm wooden grain and a single steam plume rising. Light: motivated tungsten key from the right plus soft window fill, warm 3200K. Camera: extreme macro lock-off on the pour, cut to a slow pull-back medium revealing the cup-and-saucer. Style: cinematic food commercial, shallow depth of field, appetizing color grade. Audio: ASMR cream pour, ceramic clink, no music. Constraints: no watermark, no text overlays, no Logo, no extra fingers.", "tags": [ "food", "tutorial", "macro", "official" ], "params": { "aspect": "16:9", "duration": "8-12s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "基于 Dreamina 官方 #3 - short recipe tutorial template" }, { "id": "sd-003", "model": "seedance-2.0", "category": "product-commercial", "title": "生活场景产品广告(Dreamina 官方 #11)", "prompt": "Lifestyle ad of [tech product] across multiple daily scenarios. Subject: a young user with the product. Action: shot 1 morning - the user picks the device off a nightstand and walks to the window; shot 2 office - sets it onto a desk between papers and a coffee mug; shot 3 evening - holds it on the sofa, soft smile. Scene: warm Scandinavian apartment, brand color accents kept visible across cuts. Light: natural daylight + practical warm lamps. Camera: gentle handheld with stabilizer, each shot a slow push-in to the product. Style: photorealistic lifestyle CF, consistent brand color palette across cuts. Constraints: keep brand colors and logo visible, no watermark, no random text, no double subjects, no ID drift between cuts.", "tags": [ "lifestyle", "brand", "multi-scene", "official" ], "params": { "aspect": "9:16", "duration": "12-15s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "源自 Dreamina 官方 #11 - lifestyle product across multiple daily scenarios" }, { "id": "sd-004", "model": "seedance-2.0", "category": "product-commercial", "title": "促销倒计时贴图(Dreamina 官方 #12)", "prompt": "Promotional offer with animated countdown. Subject: the brand product centered on a vibrant brand-color backdrop. Action: large animated digits tick down 5, 4, 3, 2, 1, then a clean reveal shot of the product. Scene: minimal studio set, glossy floor, soft floor reflection. Light: hard key from above-front, color-matched practical accent in brand hue. Camera: locked centered medium close-up, single slow push at the final beat. Style: high-energy retail commercial, crisp graphic feel. Audio: clean clock tick per number, single soft impact on the reveal, no voice. Constraints: avoid generating Chinese or Latin subtitle bars, only the deliberate countdown digits should appear, no watermark, no extra Logos.", "tags": [ "promo", "countdown", "retail", "official" ], "params": { "aspect": "9:16", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 #12 - promotional offer with animated countdown" }, { "id": "sd-005", "model": "seedance-2.0", "category": "product-commercial", "title": "奢华香水广告多镜头(Imagine.art 长版)", "prompt": "Ultra-luxury perfume commercial with a dreamy electronic soundtrack and steady drum beats. Shot 1 (0-5s): macro shot of a transparent rectangular glass bottle surrounded by violently swirling purple liquid, with crisp water sounds and bubbles. Shot 2 (5-10s): dissolve into surface ripples of the purple liquid, then a close-up of delicate iris flowers floating underwater. Shot 3 (10-15s): cut to the perfume bottle tilted on a textured light surface, refracting dreamy halos; a Latina female model elegantly lifts the bottle to her neck and shoulder in a pure white high-end studio, gazing into the camera. Style: high-end editorial commercial, anamorphic, shallow depth. Constraints: no watermark, no text, no Logo, no warped bottle, no extra fingers.", "tags": [ "perfume", "luxury", "multi-shot", "macro", "underwater" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 长版多镜头香水广告 (Imagine.art)" }, { "id": "sd-006", "model": "seedance-2.0", "category": "product-commercial", "title": "可乐互动叙事(ByteDance Seed 官方 I2V 范例)", "prompt": "The character in the painting looks guilty, darting glances left and right before leaning out of the picture frame. Very swiftly, they extend a hand outside the frame, grab a cola, take a sip, and reveal a look of deep satisfaction. Hearing footsteps approach, the character hastily puts the cola back just as a Western cowboy walks by and picks it up. The ending shot pushes in on a top-lit close-up of the cola against a pure black background, with artistic subtitles and a voiceover appearing at the bottom: 'Yikou Cola, an absolute must-try!' Constraints: keep character identity stable, no watermark beyond the deliberate caption.", "tags": [ "product", "narrative", "humor", "i2v", "official" ], "params": { "aspect": "16:9", "duration": "12-15s", "resolution": "1080p" }, "source": { "name": "ByteDance Seed - Seedance 2.0 official launch blog (I2V)", "url": "https://seed.bytedance.com/en/blog/official-launch-of-seedance-2-0", "tier": "official" }, "notes": "ByteDance Seed 官方博客 I2V 范例(Yikou Cola 叙事),原文一字不改" }, { "id": "sd-007", "model": "seedance-2.0", "category": "product-commercial", "title": "草图到 3D 转换(创意品牌片)", "prompt": "A car sketch on paper. The camera slowly pushes in. Pencil lines lift off the paper, gaining dimensionality, color, and reflective paint, transforming into a photorealistic 3D sports car driving on an open coastal highway. Scene: late afternoon golden light over a winding cliff road. Light: warm rim from the sun, cool fill from the sea side. Camera: a single continuous push-in that becomes a low tracking shot alongside the car after the transformation. Style: cinematic premium auto commercial, photorealistic CGI grade. Audio: pencil-on-paper scratch transitioning into a deep V8 hum. Constraints: stable car identity, no watermark, no Logo other than the sketched mark, no warped wheels.", "tags": [ "transformation", "automotive", "sketch-to-3d" ], "params": { "aspect": "16:9", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 草图变 3D 汽车(Imagine.art 原版扩写为 8 要素)" }, { "id": "sd-008", "model": "seedance-2.0", "category": "product-commercial", "title": "好莱坞科幻多分镜 10s(Volcano Engine 官方 timecode 范式)", "prompt": "Create a 10-second cinematic sequence in a Hollywood sci-fi style with cyberpunk aesthetics, high-contrast neon lighting, and an epic score feel.\n\nShot 1 (00:00-00:04) Wide tracking: a futuristic megacity canyon at night with falling rain. An anti-gravity vehicle weaves rapidly between skyscrapers. Countless holographic ads flash; neon reflections shimmer on glass curtain walls and the vehicle's surface. The vehicle leaves long blue light trails behind it.\n\nShot 2 (00:04-00:07) Cockpit tight action: the vehicle drifts sharply to avoid an obstacle. Sparks and mist as the fuselage grazes a building edge. Rain hits the camera lens, blown away by high-speed airflow. Subtle camera shake emphasizes speed.\n\nShot 3 (00:07-00:10) Epic pull-back crane: the vehicle bursts from the narrow buildings into an open city plaza. Reveal a panoramic megacity view as the vehicle flies through low clouds, with a massive lens flare streaking across the frame.\n\nTechnical: 1080p, teal-and-orange cinematic grading, HDR, complex particle effects, realistic motion blur. Constraints: no watermark, no text, no Logo, no warped buildings.", "tags": [ "sci-fi", "cyberpunk", "multi-shot", "timecode", "8-element" ], "params": { "aspect": "21:9", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 prompt guide (timecode multi-shot template)", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "依据 Volcano Engine 官方提示词指南 timecode 多分镜结构改写" }, { "id": "sd-009", "model": "seedance-2.0", "category": "social-viral", "title": "病毒夸张猫咪(Dreamina 官方 Viral #1)", "prompt": "Fast-paced viral video of a cat knocking over objects with exaggerated reactions, meme-style captions, and quick zooms for comedic effect. Subject: an orange tabby with exaggerated cartoonish eye expressions. Action: paw flicks a glass off a table; the cat's eyes go wide as the glass falls in slow motion; quick zoom on the cat's guilty face; jump cut to another object falling. Scene: a clean modern kitchen counter. Light: bright window key, slight cool tint. Camera: locked low angle on the counter, snap-zoom into the cat's face on each beat. Style: snappy short-form social, vibrant saturation. Audio: comedic sting on each impact, light boing. Constraints: no watermark, the only allowed on-screen text is the meme caption.", "tags": [ "viral", "comedy", "cat", "social", "official" ], "params": { "aspect": "9:16", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 Viral #1 - cat knocking over objects" }, { "id": "sd-010", "model": "seedance-2.0", "category": "social-viral", "title": "晨间日常 Jump-cut(Dreamina 官方 Viral #2)", "prompt": "Morning routine of a college student with upbeat background music, jump cuts between scenes, and text overlays highlighting key moments. Subject: a 20-year-old student in a hoodie. Action: alarm flick, splash water on face, pour coffee, stuff a laptop into a backpack, lock the door, step into the sunlit street. Scene: a small dorm room transitioning to a bright sidewalk. Light: warm window key in the morning, cool daylight outdoors. Camera: handheld with motion blur, each shot held about 1 second before a jump cut. Style: snappy GRWM aesthetic, slight VHS texture. Audio: upbeat lo-fi beat synced to each cut. Constraints: only the deliberate text-overlay highlights should appear, no watermark, no Logo.", "tags": [ "viral", "lifestyle", "jump-cut", "official" ], "params": { "aspect": "9:16", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 Viral #2 - morning routine" }, { "id": "sd-011", "model": "seedance-2.0", "category": "social-viral", "title": "白天到赛博朋克 VFX 转场(Dreamina 官方 #7)", "prompt": "Transform a daytime city street into a neon-illuminated cyberpunk environment with rain reflections, animated signs, and moving vehicles. Subject: a quiet urban crosswalk with a few pedestrians. Action: light fades, neon signs ignite one by one, rain begins falling, holographic ads stream across building facades, traffic light trails grow longer. Scene: a mid-rise commercial street. Light: starts mid-day overcast then transitions to high-contrast neon with magenta and cyan dominant. Camera: locked-off wide, very slow push-in throughout. Style: VFX transfer, photoreal cyberpunk, anamorphic flares. Constraints: no watermark, no extra English text on signage beyond what reads as urban decoration, no warped buildings.", "tags": [ "viral", "style-transfer", "cyberpunk", "official" ], "params": { "aspect": "9:16", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 Style & VFX Transfer #7" }, { "id": "sd-012", "model": "seedance-2.0", "category": "social-viral", "title": "森林转奇幻发光场景(Dreamina 官方 #9)", "prompt": "Convert a forest animation into a magical fantasy scene with glowing plants, floating lights, and mystical fog effects. Subject: a quiet pine forest at dusk. Action: pine needles begin to softly glow turquoise; tiny luminescent particles drift upward; ground-level mist rolls slowly between the trunks; a faint will-o-wisp orbits a tree at mid-depth. Scene: misty Pacific Northwest forest. Light: dimming twilight base layered with cool magical bioluminescence. Camera: slow gimbal forward dolly, treading at human pace. Style: magical realism, cinematic depth. Constraints: no watermark, no text, no human figure, no warped trees.", "tags": [ "viral", "fantasy", "style-transfer", "official" ], "params": { "aspect": "9:16", "duration": "10-15s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 Style & VFX Transfer #9" }, { "id": "sd-013", "model": "seedance-2.0", "category": "cinematic", "title": "短片黑屋开场(Dreamina 官方 Film #15)", "prompt": "Cinematic intro for a short film: a character enters a dimly lit room, dramatic camera pans, suspenseful music. Subject: a man in his 40s, dark overcoat, weathered face. Action: he pushes a heavy wooden door, steps in slowly, scans the room, stops in the center, exhales. Scene: a derelict study lit by a single hanging bulb over a desk; shadows pool in the corners. Light: hard practical from the overhead bulb, deep ink shadows everywhere else. Camera: slow horizontal pan tracking with him from door to desk, then a soft push-in to medium close-up. Style: noir thriller, 35mm grain, desaturated palette. Audio: low rumbling drone, distant clock tick, no dialogue. Constraints: maintain face consistency, no watermark, no subtitles.", "tags": [ "cinematic", "intro", "noir", "official" ], "params": { "aspect": "16:9", "duration": "12-15s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 Film & Game #15 - cinematic short film intro" }, { "id": "sd-014", "model": "seedance-2.0", "category": "cinematic", "title": "情感特写(银翼杀手灯光)", "prompt": "Subject: a woman's face in profile, mid-30s. Action: a single tear forms in the corner of her eye, catches the light, slowly traces down her cheek; she blinks once. Scene: a small interior, undefined background. Light: Blade Runner-inspired amber rim from camera right, teal fill from camera left, hard ratio. Camera: extreme close-up, locked-off on a tripod, no shake. Style: cinematic noir, 35mm anamorphic, photorealistic skin texture, shallow depth of field. Audio: low room tone, soft breath. Constraints: maintain face consistency, no watermark, no text, no extra ID drift.", "tags": [ "emotion", "close-up", "blade-runner", "static-camera" ], "params": { "aspect": "16:9", "duration": "8-10s", "resolution": "1080p" }, "source": { "name": "Atlas Cloud - Seedance 2.0 prompts guide", "url": "https://www.atlascloud.ai/blog/guides/best-seedance-2-0-prompts-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 单主体情感特写(Atlas Cloud 原稿 + 8 要素重写)" }, { "id": "sd-015", "model": "seedance-2.0", "category": "cinematic", "title": "1920 年代查尔斯顿舞(ByteDance Seed 官方 T2V)", "prompt": "1920s jazz club-style Charleston dance. A female dancer in a gold fringe dress and a male dancer in a striped suit perform a high-intensity routine. Moves include hyper-fast syncopated footwork, aerial catches, and exaggerated arm swings. The camera uses dynamic tracking, interspersed with close-ups of footwork. Emphasis is on physical details: the fringe swinging wildly with every kick, the sheen of sweat on their skin, and the smoky, vintage film grain cinematic texture. A background jazz band and cheering audience amplify the frenzied party atmosphere. Constraints: no watermark, no text, no Logo, characters keep consistent identity.", "tags": [ "dance", "vintage", "official", "audio-driven" ], "params": { "aspect": "16:9", "duration": "12-15s", "resolution": "1080p" }, "source": { "name": "ByteDance Seed - Seedance 2.0 official launch blog (T2V #6)", "url": "https://seed.bytedance.com/en/blog/official-launch-of-seedance-2-0", "tier": "official" }, "notes": "ByteDance Seed 官方 T2V 范例 #6(Charleston dance),原文一字不改" }, { "id": "sd-016", "model": "seedance-2.0", "category": "cinematic", "title": "雨夜霓虹侦探(赛博朋克黑色电影)", "prompt": "Subject: a detective in his 40s in a long charcoal coat. Action: stands beneath a flickering neon sign in a rain-soaked alley, slowly lights a cigarette, looks toward the camera once, takes a drag, exhales. Scene: 1940s-meets-cyberpunk alley, wet cobblestones with neon reflections rippling at every step. Light: hard neon key in magenta and cyan, inky shadow fill, single overhead practical flickering. Camera: handheld eye-level medium close-up, slight breathing motion, no whip pan. Style: cinematic noir with cyberpunk grade, 21:9 anamorphic. Audio: rain on pavement, faint distant siren, the soft hiss of the burning cigarette. Constraints: face stable, no watermark, no text, no Logo on the signage, no extra limbs.", "tags": [ "noir", "detective", "rain", "neon", "cyberpunk" ], "params": { "aspect": "21:9", "duration": "10-15s", "resolution": "1080p" }, "source": { "name": "Atlas Cloud - Seedance 2.0 prompts guide", "url": "https://www.atlascloud.ai/blog/guides/best-seedance-2-0-prompts-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 雨夜侦探 noir 单镜头 (Atlas Cloud 原稿 + 8 要素整理)" }, { "id": "sd-017", "model": "seedance-2.0", "category": "cinematic", "title": "披萨店 oner 长镜头(叙事一镜到底)", "prompt": "Single continuous cinematic shot, no music. Subject: a bearded baker in a white apron and a returning customer at the counter. Action: from outside the glass window the camera slowly pushes inward into a pizza shop. The baker removes a hot pizza from a wood oven with a metal peel, slides it into a red takeaway box, closes the lid, then hands the box across the counter to the customer with a warm nod. Scene: a small dimly lit independent pizzeria at night. Light: warm tungsten key from the oven, ambient orange wall sconces, blue cool from outside. Camera: a single continuous Steadicam push-in becoming an over-the-shoulder finish behind the customer. Style: documentary cinematic, photorealistic. Audio: oven crackle, dough slap, paper folding, no music. Constraints: maintain both faces, no watermark, no Logo, no jump cuts.", "tags": [ "oner", "long-take", "documentary", "warm" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 披萨店 oner (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-018", "model": "seedance-2.0", "category": "cinematic", "title": "女子排球比赛(多机位)", "prompt": "A women's volleyball match in a spacious indoor gym with an arched wooden ceiling and large windows. Subject: a blue team and a green team plus a middle-aged female coach. Action: the blue and green teams exchange intense rallies while players shout to communicate; the green team fails to save the ball, an off-screen whistle blows. Cut to a close-up of the coach with short blonde hair, metal-framed glasses, and a green jacket - she watches intently, then lowers her head in frustration. Scene: a packed-stand gym with colorful banners. Light: high-key gym daylight from the side windows. Camera: medium-wide on the rally, then a sharp cut to the coach close-up. Style: photoreal sports documentary. Audio: shouts, ball impact, whistle, energetic rhythmic background track. Constraints: keep coach identity stable across cuts, no watermark, no scoreboard text overlays.", "tags": [ "sports", "drama", "multi-character", "audio" ], "params": { "aspect": "16:9", "duration": "12-15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 室内排球比赛 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-019", "model": "seedance-2.0", "category": "cinematic", "title": "夜街舞 battle 15s(多分镜)", "prompt": "Generate a 15-second photorealistic street dance battle at night.\n\nShot 1 (0-5s): one dancer performs challenging solo floor moves in the center of a semicircle of dancers; low tracking handheld behind the group.\n\nShot 2 (5-15s): the full group performs synchronized choreography combining breaking, popping, locking, and hip-hop; smooth dolly arc around the formation.\n\nScene: a wet urban street at night with reflective pavement, neon shop signs, a thin layer of mist. Light: neon magenta and cyan keys plus a single soft overhead practical. Camera: stable cinematic handheld throughout, no whip pans. Style: photoreal urban cinematic, anamorphic feel, realistic physics and body proportions. Audio: deep bass drop, syncopated drum beats, ambient city. Constraints: no watermark, no on-screen text, no extra limbs, no warped faces, no twins.", "tags": [ "dance", "urban", "night", "neon", "multi-character" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 prompt guide (multi-shot timecode template)", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "改写为 Volcano Engine 官方多分镜 timecode 模板(Shot 1 + Shot 2)" }, { "id": "sd-020", "model": "seedance-2.0", "category": "cinematic", "title": "洞穴探险第一视角(音频丰富)", "prompt": "First-person POV handheld footage with slight natural shake. Subject: an unseen explorer with a headlamp and visible hand occasionally entering frame. Action: the camera moves through a narrow tunnel; the hand enters frame to brace against the wall; the explorer pauses, then pans slowly left to right over a chamber of glowing semi-transparent crystals. Scene: a damp mossy cave tunnel opening into a crystal cavern. Light: hard headlamp beam as the only key, deep black surroundings. Camera: POV handheld with realistic micro-shake. Style: photoreal documentary horror tension. Audio: heavy breathing, footsteps on wet soil, sharp footfall clacks echoing in the cavern, an awed exhale that swirls and fades, no music. Constraints: no watermark, no on-screen text, no extra hands.", "tags": [ "pov", "exploration", "atmospheric", "audio-rich" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 洞穴 POV (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-021", "model": "seedance-2.0", "category": "cinematic", "title": "电影分镜直转视频(@storyboard 蓝图)", "prompt": "Use @[storyboard ref] as the authoritative 15-second cinematic shot blueprint. Do not render the storyboard sheet itself: exclude all panel borders, headers, text, labels, project card, director. Subject: faithfully reproduce only the in-shot content drawn in each panel. Action: follow shot order, scale and camera direction exactly as drawn. Scene, light, camera: match the storyboard frame composition; respect the indicated lens and angle for each panel. Style: cinematic continuity, no extra inserts. Constraints: no watermark, no subtitles, no project card, no director name, no panel borders bleeding into the render.", "tags": [ "reference", "storyboard", "cinematic", "multi-shot" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 prompt guide (@-reference storyboard pattern)", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "Volcano Engine 官方 @ 参考蓝图模板:把整张分镜图作为权威参考" }, { "id": "sd-022", "model": "seedance-2.0", "category": "cinematic", "title": "钢琴家蝴蝶光(中景情感)", "prompt": "Subject: a pianist in his 30s wearing a simple, elegant black suit. Action: his fingers move across the keys with precision and emotion; he leans slightly into the keyboard during a crescendo, then settles back. Scene: a concert hall stage at evening. Light: classic butterfly lighting from above-front (slightly higher than 45 degrees), a single warm spotlight, soft fall-off behind. Camera: medium shot, locked-off, no movement. Style: cinematic concert film, shallow depth of field. Audio: solo piano, faint audience presence in the background. Constraints: maintain face stable, no watermark, no text, no extra hands.", "tags": [ "music", "portrait", "butterfly-light", "medium-shot" ], "params": { "aspect": "16:9", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 钢琴家蝴蝶光 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-023", "model": "seedance-2.0", "category": "cinematic", "title": "稳定器长镜头:咖啡馆-街道-地铁", "prompt": "Steadicam long take. Subject: a man in his 30s in a navy jacket. Action: in a warm yellow-lit cafe he pushes the door open, walks through traffic with car headlights creating long light trails on wet asphalt, then descends a dim subway staircase. Scene: warm interior cafe to cool-toned urban street to fluorescent subway. Light: the lighting transitions smoothly from warm 3000K to neutral daylight to cool 5600K subway fluorescents, all motivated. Camera: a single continuous Steadicam follow, never cutting. Style: cinematic urban documentary, anamorphic feel. Audio: cafe chatter and door bell, traffic, then echoey subway hum. Constraints: maintain subject identity, no watermark, no text, no jump cuts.", "tags": [ "steadicam", "long-take", "light-transition", "urban" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 咖啡馆到地铁稳定器长镜头 (Imagine.art 原稿)" }, { "id": "sd-024", "model": "seedance-2.0", "category": "cinematic", "title": "王家卫式港片复古", "prompt": "Subject: a person in a khaki trench coat, gender intentionally ambiguous, at a rain-soaked red phone booth. Action: listens silently with the receiver held to one ear, lips trembling slightly with restrained emotion; hangs up and walks into the rainy night crowd. Scene: 1990s Hong Kong street at night, neon signs reflecting in puddles. Light: yellow-green tinted neon ambient, hard rim from a nearby window, low key. Camera: handheld extreme close-up to medium, simulated step-printing for the walk-away. Style: 90s Hong Kong art cinema with retro grain, step-printing effect, motion-blur smear, shallow depth of field. Audio: rain on glass, ambient street murmur, no music. Constraints: maintain face stable, no watermark, no Chinese subtitles, no extra limbs.", "tags": [ "wong-kar-wai", "vintage", "hk-cinema", "step-printing" ], "params": { "aspect": "21:9", "duration": "12-15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 王家卫港片风格 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-025", "model": "seedance-2.0", "category": "action", "title": "@图1+@图2 角色 + 绿幕动作迁移格斗(R2V)", "prompt": "Make the character from @Image1 and the character from @Image2 fight using the movements from the green-screen reference @Video1. Subject: two clearly distinct fighters, with face features faithfully taken from each image. Scene: abandoned multi-level parking lot at dusk. Action: choreography matches @Video1 frame-for-frame, with realistic impact reactions on each contact. Light: practical sodium-vapor lamps casting long shadows, cool fill from outside. Camera: dynamic gimbal arc tracking the fight, no whip pans. Style: photoreal action cinematic, realistic impact physics. Audio: impact thuds, fabric whoosh, no dialogue. Constraints: maintain each fighter's face and outfit; do NOT swap identities; no watermark, no text, no Logo, no twins.", "tags": [ "combat", "r2v", "green-screen", "reference", "official" ], "params": { "aspect": "16:9", "duration": "10-15s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 @-reference R2V template", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "采用 Volcano Engine 官方 @ 参考语法(@Image1/@Image2/@Video1)改写" }, { "id": "sd-026", "model": "seedance-2.0", "category": "action", "title": "雪地近身格斗(女特工)", "prompt": "Subject: a short-haired female agent in tactical winter gear and a mercenary in dark armor. Action: the mercenary throws a punch; she dodges swiftly and counters with a powerful elbow strike to his face mask, followed by a heavy knee to the stomach; impact shows visible shockwave ripples, sweat and saliva flying as his body folds from the blow. Scene: a snowy military base perimeter at dawn. Light: cold blue-white ambient with hard backlight, vapor breath catching the light. Camera: dynamic handheld tracking the contact moments, shallow shake on impact. Style: gritty photoreal action thriller, high contrast. Audio: heavy breathing, impact thuds, snow crunch, no music. Constraints: maintain both faces, no watermark, no text overlays, no extra limbs, no twins.", "tags": [ "combat", "winter", "handheld", "gritty" ], "params": { "aspect": "16:9", "duration": "12s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 雪地近身格斗 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-027", "model": "seedance-2.0", "category": "action", "title": "北欧雪夜人狼对峙", "prompt": "Realistic Nordic snowy night sequence.\n\nShot 1 (0-5s): a man runs fast through deep snow, tracked closely from behind at foot level.\n\nShot 2 (5-9s): a wolf chases him; the man trips and rolls; the wolf lunges.\n\nShot 3 (9-12s): a border collie leaps in, knocking the wolf aside; the animals wrestle; the wolf breaks free and howls.\n\nShot 4 (12-15s): multiple wolves slowly surround the man and dog; close-up on the dog as the man slowly sits up.\n\nLight: cold moonlight, faint warm rim from a distant cabin window. Camera: low handheld tracking on the runs, medium static on the wrestle, slow push-in for the close-up. Style: photoreal survival thriller. Audio: snow crunch, panting, low growl, escalating tense score. Constraints: maintain the man's identity, no watermark, no on-screen text, no extra limbs.", "tags": [ "chase", "winter", "animals", "tense" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 prompt guide (multi-shot timecode template)", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "改写为 Volcano Engine 官方多分镜 timecode 模板(四镜头叙事)" }, { "id": "sd-028", "model": "seedance-2.0", "category": "action", "title": "赛博朋克刺客(游戏 CG)", "prompt": "Subject: a young assassin in dark armor with a glowing energy lightsaber. Action: the assassin fights multiple masked enemies; the camera rapidly pulls back to reveal them surrounding him; he moves fluidly, defeating enemies one by one in clean choreographed beats; the sequence stays fast but smooth; teleportation-style afterimages blend with cinematic transitions; he ends by looking directly at the camera. Scene: a dark cyberpunk city corner with holographic ads and rain. Light: harsh neon magenta and cyan, hard rim. Camera: starts close on the assassin then arcs out to reveal full surroundings, then snaps back to close-up on the final stare. Style: game cinematic CG, photoreal cyberpunk. Audio: synth bass, hum of the energy blade, impact thuds. Constraints: maintain assassin identity, no watermark, no text, no extra limbs.", "tags": [ "cyberpunk", "combat", "game-cgi", "vfx" ], "params": { "aspect": "16:9", "duration": "12-15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 赛博朋克刺客 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-029", "model": "seedance-2.0", "category": "action", "title": "涩谷跑酷一镜到底(多音轨)", "prompt": "Single continuous cinematic shot in Shibuya. Subject: a Japanese schoolgirl in uniform, hair tied back. Action: she sprints at extreme speed, weaves along crosswalks, runs along low walls, leaps, rolls on landing, dashes forward again. Scene: Shibuya scramble crossing at dusk, neon signs above, busy crowd. Light: warm sunset key plus emerging neon practicals, motivated. Camera: low and close to the ground, following every leap and roll; gimbal stabilized, no whip pans. Style: photoreal urban parkour cinematic. Audio mix: daytime street noise, retro 8-bit game sounds layered on jumps, shamisen plucks accenting beats. Constraints: maintain face stable, no watermark, no on-screen text, no doubles.", "tags": [ "parkour", "tokyo", "oner", "low-angle" ], "params": { "aspect": "9:16", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 涩谷跑酷一镜到底 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-030", "model": "seedance-2.0", "category": "action", "title": "霓虹雨夜城市追逐", "prompt": "High-energy cinematic chase at night in a neon-lit city. Subject: a lone runner in a dark jacket. Action: sprints through rain-soaked streets; police drones and headlights blur past; close-ups of eyes, boots splashing puddles, wide traffic shots intercut; vaults a barrier, slides across a car hood, dodges a small explosion; ends in slow motion leaping off a rooftop with city lights streaking beneath. Scene: dense downtown at night, wet pavement everywhere. Light: neon magenta and cyan keys, hard practical sodium, harsh emergency strobes on cuts. Camera: handheld style with motion blur, dynamic angle changes, no double-camera moves in one shot. Style: photoreal action cinematic, anamorphic feel. Audio: pounding bass, sirens, rain. Constraints: maintain face stable, no watermark, no text, no extra limbs.", "tags": [ "chase", "neon", "rain", "handheld" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 霓虹城市追逐 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-031", "model": "seedance-2.0", "category": "action", "title": "城市追逐分镜(Dreamina 官方 Film #13)", "prompt": "Storyboard a chase scene in a busy city with multiple camera angles, dynamic character movements, and realistic environmental interactions. Subject: a man in a red hoodie being chased by a man in a black coat. Action: weave through pedestrians, leap a low railing, slide across a parked car hood, knock over a fruit cart, climb a fire escape; the pursuer mirrors each move with a 1-second delay. Scene: a busy midtown street at lunchtime. Light: bright midday with hard shadow ratio. Camera: low handheld tracking, alternating with high-angle overhead inserts, no whip pans. Style: cinematic urban thriller, realistic physics. Audio: footsteps, shouts, traffic, racing pulse score. Constraints: maintain both faces, no watermark, no text, no extra limbs.", "tags": [ "chase", "urban", "storyboard", "official" ], "params": { "aspect": "16:9", "duration": "12s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 Film & Game Previz #13 - chase storyboard" }, { "id": "sd-032", "model": "seedance-2.0", "category": "action", "title": "@图1 战场战术射击(R2V)", "prompt": "The character defined by @Image1 as the protagonist, agile and athletic. Action: he runs through a battlefield under a hail of bullets while engaging in intense tactical shooting; rolls behind a sandbag, takes aim, fires three precise shots, reloads, sprints forward again. Scene: an open warzone with smoke pillars and twisted debris. Light: harsh midday sun cut by smoke, deep contrast. Style: cinematic with strong teal-and-orange grading, high contrast, visible film grain. Camera: handheld tracking shaking violently alongside the protagonist with occasional snap zooms for extreme tension. Audio: distinct muzzle flashes paired with sharp gunshots, bullets striking surrounding objects creating flying rubble and sparks, shockwaves and smoke from explosions in the background. Character performance: determined eyes, actions filled with power and speed, replicating a Hollywood action superstar. Constraints: maintain face from @Image1, no watermark, no Logo, no extra limbs.", "tags": [ "military", "tactical", "teal-orange", "handheld", "r2v" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 prompt guide (@-reference R2V pattern)", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "改写为 Volcano Engine 官方 @-reference R2V 模板(@Image1 主角锁定)" }, { "id": "sd-033", "model": "seedance-2.0", "category": "sports", "title": "花样滑冰双人赛(ByteDance Seed 官方 T2V)", "prompt": "Competitive figure skating live performance. Opening with a low-angle shot following the skate blades gliding on ice, with clear details of ice shavings and light reflections. Entering the spin sequence: the male skater makes a minor error due to slight axis deviation, causing a brief disruption in spin rhythm. The female skater quickly adjusts her center of gravity, keeps a calm gaze, and signals 'Stay with me' with eye contact, guiding the male skater to regain alignment. This transitions seamlessly into clean, stable lift movements with sharp, consistent body posture. The climax is a synchronized jump combination: upright aerial posture, decisive ice landings, with perfect audio-visual alignment. The female skater wears a deep-blue figure skating dress; the male skater is in competitive athletic attire. The full sequence delivers a complete narrative from tense mistake to calm, successful execution, highlighting the technical skill and mental strength of elite pair figure skating. Constraints: no watermark, no text overlays, maintain both skater identities.", "tags": [ "sports", "figure-skating", "low-angle", "official", "synchronized" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "ByteDance Seed - Seedance 2.0 official launch blog (T2V #1)", "url": "https://seed.bytedance.com/en/blog/official-launch-of-seedance-2-0", "tier": "official" }, "notes": "ByteDance Seed 官方 T2V 范例 #1 (figure skating),原文一字不改" }, { "id": "sd-034", "model": "seedance-2.0", "category": "sports", "title": "好莱坞夜雨赛车 timecode(多分镜)", "prompt": "Hollywood professional racing style at night in the rain.\n\nShot 1 (00-05s): close-up inside the veteran driver's car as rain lashes the windshield. He stays calm, nods, and mouths 'Let's go.'\n\nShot 2 (05-10s): close-up inside the rival car; the younger driver grips the wheel tightly, breathing heavily, whispering 'Focus.'\n\nShot 3 (10-15s): wide action shot as the green lights trigger both cars to accelerate on wet asphalt, spraying water into the camera; motion blur turns stadium lights into streaks of color.\n\nLight: stadium overheads through rain, harsh headlight key, deep shadow. Camera: locked interior close-ups for shots 1-2, then a low exterior track for shot 3. Style: photoreal motorsport cinematic, anamorphic flares. Audio: rain on windshield, engine idle then roar, tire spray. Constraints: maintain both driver identities, no watermark, no Logo on vehicles other than the deliberate brand mark, no extra hands.", "tags": [ "racing", "rain", "multi-shot", "hollywood" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 prompt guide (multi-shot timecode template)", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "改写为 Volcano Engine 官方 timecode 多分镜模板(三镜头 5s + 5s + 5s)" }, { "id": "sd-035", "model": "seedance-2.0", "category": "sports", "title": "武侠对决:白衣 vs 黑衣(三分镜)", "prompt": "Eastern Wuxia-style duel between a grandmaster in white and a grandmaster in black on an ancient stone platform.\n\nShot 1 (0-5s): both clash at high speed, leaving afterimages; sparks and a ring-shaped shockwave erupt; the flagstone cracks, debris flies.\n\nShot 2 (5-10s): they soar into the air, striking amid falling boulders and shattered pillars, with dust and air distortion emphasizing destructive force.\n\nShot 3 (10-15s): both release ultimate moves, cyan and red energy orbs collide, triggering an earth-shattering explosion.\n\nLight: stormy overcast, hard rim from cracked sun beams. Camera: stable cinematic with controlled push-ins on the strikes, no whip pans. Style: photoreal Wuxia cinematic, anamorphic, epic score feel. Audio: thunder roll, blade clash, explosion bass. Constraints: maintain both characters, no watermark, no text, no extra limbs, no twins.", "tags": [ "wuxia", "duel", "vfx", "epic" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 prompt guide (multi-shot timecode template)", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "改写为 Volcano Engine 官方多分镜 timecode 模板(武侠三镜头)" }, { "id": "sd-036", "model": "seedance-2.0", "category": "sports", "title": "竹林雨中剑客(ByteDance Seed 官方 T2V)", "prompt": "A Wuxia-style audiovisual blockbuster. A white-clad swordsman and a straw-caped blademaster face off in a bamboo forest. The camera slowly pushes in between them, shifting focus between raindrops and sword hilts. The atmosphere is extremely oppressive; only the sound of rain can be heard. Suddenly, a crack of thunder flashes, and both charge simultaneously. A fast-panning profile shot captures their mud-splattering footsteps. The precise moment their weapons clash, the footage switches to ultra-slow motion, clearly displaying the ring-shaped shockwaves of rainwater blasted away by the blades, along with bamboo leaves sliced by sword aura. Normal speed resumes as they land back-to-back. The straw-caped blademaster's bamboo hat splits open, and the scene cuts abruptly. Constraints: no watermark, no on-screen text, maintain both swordsmen identities.", "tags": [ "wuxia", "rain", "bamboo", "official", "slow-motion" ], "params": { "aspect": "16:9", "duration": "12-15s", "resolution": "1080p" }, "source": { "name": "ByteDance Seed - Seedance 2.0 official launch blog (T2V #4)", "url": "https://seed.bytedance.com/en/blog/official-launch-of-seedance-2-0", "tier": "official" }, "notes": "ByteDance Seed 官方 T2V 范例 #4 (Wuxia bamboo forest duel),原文一字不改" }, { "id": "sd-037", "model": "seedance-2.0", "category": "sports", "title": "风筝跑酷长镜头(风格切换)", "prompt": "Subject: a Chinese youth in a light jacket flying a long-tailed kite. Action: in a single continuous shot he runs through crowded narrow streets, leaps up a flight of stone steps, performs a controlled mid-air flip, lands and immediately dashes toward a high platform for a difficult jump; a drumbeat erupts at landing. Scene: a dense old-town alley opening into a wide plaza. Light: 0-5s warm saturated daylight (Makoto Shinkai style with moist reflections), 6-10s shifts to ink-wash monochrome with simplified buildings and visible speed lines. Camera: a single steadicam follow throughout. Style: hybrid Makoto Shinkai then ink-wash with ink shockwaves visualizing the drumbeats. Audio: pace breath, kite paper, drum kick on landing. Constraints: maintain face stable across the style transition, no watermark, no text, no extra limbs.", "tags": [ "parkour", "oner", "style-transition", "shinkai", "ink-wash" ], "params": { "aspect": "9:16", "duration": "10s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 风筝跑酷长镜头 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-038", "model": "seedance-2.0", "category": "nature", "title": "深海生物发光潜水(POV)", "prompt": "Subject: an unseen scuba diver wearing a torch. Action: descends into a pitch-black ocean; the torch beam cuts a narrow cone; at depth she clicks the torch off; total darkness for two seconds; faint blue pulses appear; bioluminescent creatures begin appearing in the void - a jellyfish drifts past trailing light, a school of fish materializes as a glowing spiral; the diver's hand reaches out slowly; the fish scatter like sparks; the camera pans across the dark to reveal a vast structure of glowing coral below. Scene: deep open water transitioning to a coral structure. Light: only the torch and the bioluminescence, otherwise pitch black. Camera: first-person POV, gentle controlled motion. Style: photoreal nature doc, ethereal. Audio: slow exhale on the soundtrack, faint underwater whoosh, no music. Constraints: no watermark, no text, no extra hands.", "tags": [ "pov", "underwater", "bioluminescent", "atmospheric" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 prompt guide (POV + bioluminescence audio template)", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "改写为 Volcano Engine 官方 POV + audio-driven 模板(slow exhale 声轨与画面同步,体现 dual-channel audio 能力)" }, { "id": "sd-039", "model": "seedance-2.0", "category": "nature", "title": "城市日转夜锁定延时", "prompt": "Subject: an entire dense city skyline. Action: locked-off wide shot from a high vantage; morning light sweeps across the rooftops, shadows rotate, clouds roll through in fast motion, afternoon haze settles, then city lights ignite one cluster at a time as dusk falls; the final ten seconds slow to real time with a helicopter tracking slowly across frame. Scene: a generic large city downtown. Light: full daylight cycle baked into the timelapse, ending in cool night with motivated practical city light. Camera: a single locked tripod position throughout, no cuts. Style: nature-doc timelapse cinematic. Audio: subtle ambient city drone, single helicopter blade hum at the end. Constraints: no watermark, no text overlays, no scene cuts.", "tags": [ "timelapse", "city", "day-to-night", "locked-off" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 prompt guide (lighting-first locked-off template)", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "改写为 Volcano Engine 官方'锁定机位 + 光影变化作为唯一驱动'范式(lighting 是官方指南指定的第一权重要素)" }, { "id": "sd-040", "model": "seedance-2.0", "category": "nature", "title": "龙卷风追风者车载镜头", "prompt": "Subject: a stationary dashcam inside a storm-chaser pickup truck and a fully-formed tornado outside. Action: the truck drives toward the funnel on a flat American plain; the driver's hand enters frame to adjust the camera; light rain begins then becomes heavy; wipers struggle; lightning strikes ahead; cut to a wide exterior of the small vehicle against the vast sky continuing toward the storm. Scene: a yellow-green sky over the plains, the funnel tracking left to right across the road ahead. Light: ominous sickly-green dome lit, hard lightning bursts. Camera: handheld dashcam interior shake, then a wide locked exterior on the cut. Style: photoreal storm-chasing documentary. Audio mix: raw wind, static radio chatter, rain intensity building. Constraints: no watermark, no on-screen text, no warped vehicle.", "tags": [ "storm", "dashcam", "documentary", "audio-rich" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Volcano Engine - Seedance 2.0 prompt guide (multimodal audio-driven template)", "url": "https://www.volcengine.com/docs/82379/2222480", "tier": "official" }, "notes": "改写为 Volcano Engine 官方'多模态音视频联合生成'范式(dual-channel audio + ambient sound + radio chatter)" }, { "id": "sd-041", "model": "seedance-2.0", "category": "nature", "title": "北极光延时(小屋灯光)", "prompt": "Subject: a lone cabin with a single amber window and a snow-covered plain. Action: locked timelapse - the aurora emerges, first a pale green ribbon, then sheets of green and violet ripple across the sky, folding and expanding; stars are visible behind the curtains of light; a silhouetted figure steps out of the cabin and looks up; the timelapse slows to real time for the final five seconds. Scene: Iceland or northern Norway in winter. Light: black-blue sky base, cabin window amber practical, dancing aurora. Camera: locked tripod throughout, no movement. Style: photoreal nature timelapse, cinematic. Audio: complete silence except for a low ambient wind tone. Constraints: no watermark, no text, no extra figures, no warped horizon.", "tags": [ "aurora", "timelapse", "atmospheric", "locked-off" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 北极光延时 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-042", "model": "seedance-2.0", "category": "nature", "title": "山顶日出徒步者", "prompt": "Subject: a solo hiker in a small silhouette against an enormous orange dawn sky. Action: she crests a mountain ridge, stops, sets down her pack, and slowly raises both arms; the camera holds; a very gradual push-in begins from a wide establishing shot to a medium over-the-shoulder looking out at the range below. Scene: an alpine summit at first light. Light: warm orange sun key from frame right, cool blue shadow on her near side, hard rim along the ridge. Camera: locked-off wide for the first beat, then a slow gentle push-in, no whip pans. Style: photoreal nature documentary. Audio: only wind moving through the frame, the crunch of boots on gravel settling into stillness, no music. Constraints: maintain face stable, no watermark, no text, no extra limbs.", "tags": [ "sunrise", "documentary", "hiking", "atmospheric" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 山顶日出 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-043", "model": "seedance-2.0", "category": "quiet-moments", "title": "雨夜便利店(ByteDance Seed 官方场景延伸)", "prompt": "A girl hangs laundry gracefully. After finishing, she takes another piece of clothing from the bucket and shakes it vigorously. Extension: the scene shifts to her stepping into a small Asian convenience store at midnight, rain streaks the glass behind her, warm fluorescent light floods her face, she walks slowly down the snack aisle trailing one finger along the shelves, stops, picks up a cup of instant noodles, reads the label, puts it back, then grabs it again and walks to the counter; a wide shot from outside the glass as she pays, rain blurring the frame. Light: motivated cool fluorescent indoors, cool blue rain exterior. Camera: single continuous Steadicam interior, then a static wide exterior on the cut. Style: photoreal slice-of-life cinematic. Audio: ambient store hum, rain, no music. Constraints: maintain face stable, no watermark, no on-screen text, no jump cuts.", "tags": [ "slice-of-life", "convenience-store", "rain", "oner", "official" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "ByteDance Seed - Seedance 2.0 official launch blog (T2V #2)", "url": "https://seed.bytedance.com/en/blog/official-launch-of-seedance-2-0", "tier": "official" }, "notes": "前半段为 ByteDance Seed 官方 T2V #2 原文(girl hangs laundry),后半段沿同主体延伸到便利店" }, { "id": "sd-044", "model": "seedance-2.0", "category": "quiet-moments", "title": "屋顶日出瑜伽", "prompt": "Subject: a young woman in athletic wear practicing yoga alone. Action: she moves through a sun salutation sequence with fluid deliberate transitions on a high city rooftop at sunrise. Scene: a flat tar roof with the city skyline stretching behind her, golden hour catching her silhouette. Light: warm sunrise key from camera right, cool soft fill from the sky, hard rim along her arms. Camera: a slow crane shot from below the roofline rising until the full cityscape is revealed behind her; shallow depth keeps the background a soft bokeh of warm light and steel. Style: photoreal lifestyle cinematic. Audio: ambient wind, distant city hum, no music. Constraints: maintain face stable, no watermark, no on-screen text, no extra limbs.", "tags": [ "yoga", "sunrise", "crane", "rooftop", "golden-hour" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 屋顶瑜伽 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-045", "model": "seedance-2.0", "category": "quiet-moments", "title": "深夜拉面馆 oner", "prompt": "Single continuous Steadicam shot. Subject: a salaryman at the counter and a chef behind the counter. Action: from the kitchen end the camera tracks along the counter to the customer; the chef moves with practiced economy - ladles broth, pulls noodles from the pot, places toppings; close-up: the salaryman lifts his chopsticks and takes his first bite, exhales, and his shoulders drop slightly; the camera holds. Scene: a ramen shop in Tokyo at 1am, near empty. Light: warm tungsten practicals over the counter, deep dark behind. Camera: a single continuous Steadicam push, never cutting. Style: photoreal slice-of-life cinematic. Audio: low exhaust fan hum, distant rain, faint sound of a late train, no music. Constraints: maintain both subjects' identities, no watermark, no on-screen text, no jump cuts.", "tags": [ "slice-of-life", "ramen", "steadicam", "japan", "night" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 深夜拉面馆 oner (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-046", "model": "seedance-2.0", "category": "fantasy-scifi", "title": "外星丛林探索", "prompt": "Subject: a team of three explorers in advanced exosuits with chest-mounted lights. Action: they move cautiously through a dense alien jungle; the camera pans and tilts to reveal enormous alien trees and mist rising from the ground; bioluminescent plants softly glow around them; strange creatures watch from a distance. Scene: a deep alien rainforest, gigantic plant scale. Light: native bioluminescence as primary, exosuit beams as moving practicals, cool moonlit fill. Camera: shoulder-stabilized handheld with gentle motion. Style: photoreal sci-fi cinematic. Audio: alien wildlife clicks, suit servo hums, soft breathing. Constraints: maintain explorer identities, no watermark, no text, no extra limbs.", "tags": [ "sci-fi", "alien", "exploration", "bioluminescent" ], "params": { "aspect": "16:9", "duration": "12-15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 外星丛林探索 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-047", "model": "seedance-2.0", "category": "fantasy-scifi", "title": "蒸汽朋克飞艇大战", "prompt": "Subject: a massive steampunk airship and several smaller darting airships in battle. Action: the main ship's mechanical cannons fire with bursts of steam; smaller ships dart around making strafing runs; crew members operate levers, pulleys, and brass weaponry with precision; explosions throw sparks into the sky. Scene: a sepia-toned cityscape skyline at sunset. Light: warm orange sun key, smoke catches the beam, hard ratio. Camera: angles shift dynamically - wide establishing of the main ship, then a tight handheld over the deck around the gunner, then a low arc following a strafing run. Style: photoreal steampunk cinematic. Audio: brass valve hiss, deep cannon thump, distant alarm bell. Constraints: no watermark, no text, no Logo, no warped ship hulls.", "tags": [ "steampunk", "airship", "battle", "vfx" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 蒸汽朋克飞艇 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-048", "model": "seedance-2.0", "category": "fantasy-scifi", "title": "黎明武士决斗(樱花山雾)", "prompt": "Subject: two samurai facing each other at the crest of a misty mountain at dawn. Action: the first light hits their blades as they bow, eyes locked; each movement is deliberate and slow; the wind rustles through nearby cherry blossoms; a close-up on hands gripping hilts, then a wide shot showing the mist rolling over the valley below; the duel begins with a flurry of precise strikes; sparks fly from blade clashes. Scene: an open mountain ridge, cherry blossoms in the middle ground. Light: cold blue dawn key, warm rim from the rising sun. Camera: locked wide for the bow, then a tight tracking circle around the strike. Style: photoreal samurai cinematic, anamorphic feel. Audio: wind, sword draw, single shakuhachi note. Constraints: maintain both samurai identities, no watermark, no text, no extra limbs.", "tags": [ "samurai", "duel", "mountain", "cherry-blossom" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 黎明武士决斗 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-049", "model": "seedance-2.0", "category": "fantasy-scifi", "title": "奇幻战斗可视化(Dreamina 官方 #14)", "prompt": "Visualize a fantasy battle between heroes and monsters in a forest with magic effects, detailed terrain, and animated camera sweeps. Subject: three heroes in armor versus two large monstrous creatures. Action: the heroes cast bright spells with cyan and gold particle bursts; the monsters lunge across uneven forest terrain; the heroes evade and counterattack; the camera sweeps from a high arc down to ground level on a critical strike. Scene: an old-growth fantasy forest with shafts of light through the canopy. Light: motivated god-rays through trees, magical practicals on spells, soft cool fill. Camera: arc-down sweep, then a low tracking on the climactic strike, no whip pans. Style: photoreal fantasy cinematic, AAA-game adjacent. Audio: spell whoosh, monster roar, impact thud, orchestral hits. Constraints: maintain hero identities, no watermark, no text, no extra limbs, no twin monsters.", "tags": [ "fantasy", "battle", "vfx", "official" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 Film & Game #14 - fantasy battle visualization" }, { "id": "sd-050", "model": "seedance-2.0", "category": "fantasy-scifi", "title": "巨型机甲对战巨龙", "prompt": "Subject: a towering humanoid mech and a massive scaled dragon. Action: the mech fires charged energy beams from its forearms; the dragon dodges with agile aerial flips, then closes for a clawed strike; the mech twists at the waist to block, sparks fly; lightning crackles as their powers clash. Scene: a crumbling city, columns of smoke rising. Light: stormy overcast key, harsh lightning bursts on every clash. Camera: angles switch between wide establishing of the duel, intense close-ups of the mech's targeting visor and the dragon's eyes, and a low hero shot on the final clash. Style: photoreal blockbuster cinematic, anamorphic feel. Audio: deep mech servo, dragon roar, lightning crack, orchestral hit. Constraints: no watermark, no text, no Logo, no extra limbs, no twin dragons.", "tags": [ "mech", "dragon", "battle", "epic" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 机甲对决巨龙 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-051", "model": "seedance-2.0", "category": "creative", "title": "AR 购物互动设计(Dreamina 官方 #17)", "prompt": "Design an AR shopping experience where users click on products, view details, and see animated previews with realistic shadows. Subject: a young user holding a phone, viewing AR product cards floating above a coffee table. Action: she taps a sneaker card; the card lifts and rotates to reveal product details; the sneaker animates in mid-air with a realistic shadow on the table; she swipes to the next card; another product appears. Scene: a bright modern living room. Light: warm window key, even soft ambient. Camera: locked-off over-the-shoulder framing on the phone screen with a slow push-in to the floating product. Style: AR commercial cinematic, high-key, slight glow on cards. Audio: soft tap, smooth UI swoosh, no music. Constraints: maintain user identity, no watermark, no Logo other than the deliberate brand mark on the products.", "tags": [ "ar", "interactive", "ux", "creative", "official" ], "params": { "aspect": "9:16", "duration": "12s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 Creative & Interactive #17" }, { "id": "sd-052", "model": "seedance-2.0", "category": "creative", "title": "互动叙事多分支(Dreamina 官方 #18)", "prompt": "Interactive storytelling video where viewers make choices, and characters respond differently depending on the path selected. Subject: a young protagonist in a fantasy village square facing two NPCs. Action: shot 1 - protagonist arrives, two branching choice tiles fade in; shot 2 (path A) - she approaches the merchant who hands her a glowing key; shot 3 (path B) - she approaches the warrior who hands her a sword; ending - she walks toward a portal with the chosen item glowing in her hand. Scene: a stylized fantasy village. Light: warm afternoon key, magical practical glow on the items. Camera: locked-off medium for choice moments, slow push-in for each path. Style: photoreal cinematic with subtle stylization. Audio: gentle ambient lute, soft chime on each choice. Constraints: maintain protagonist identity across all branches, no watermark, no text other than the deliberate choice tiles.", "tags": [ "interactive", "narrative", "fantasy", "official" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 Creative & Interactive #18" }, { "id": "sd-053", "model": "seedance-2.0", "category": "creative", "title": "@图1 超级英雄 IP 多场景(Dreamina 官方 #4)", "prompt": "Animate a superhero performing a signature move across different city rooftops while keeping costume, hairstyle, and facial features consistent. Subject: the superhero defined by @Image1. Action: shot 1 - lands a heroic three-point pose on a glass skyscraper rooftop; shot 2 - leaps to the next rooftop, cape flowing; shot 3 - lands and fires a signature beam from his palms. Scene: three distinct city rooftops, sunset to night. Light: warm sunset key in shot 1, mixed neon-and-blue in shot 3; consistent rim across cuts. Camera: low hero angle in shot 1, dynamic side tracking in shot 2, slow push-in for the signature in shot 3. Style: photoreal blockbuster cinematic. Audio: cape whoosh, impact thuds, low orchestral pulse. Constraints: keep costume, hairstyle and face from @Image1 identical across all cuts, no watermark, no text, no doubles.", "tags": [ "ip", "character-consistency", "superhero", "official" ], "params": { "aspect": "16:9", "duration": "12s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "Dreamina 官方 Character & IP Consistency #4" }, { "id": "sd-054", "model": "seedance-2.0", "category": "asmr", "title": "护肤品 ASMR 特写(竖屏)", "prompt": "Vertical ASMR video with no music, focusing on macro details. Subject: a light-blue skincare gel bottle on glass and a pale elegant hand. Action: the hand gently taps the glass producing crisp fingernail clicks; picks up the bottle and slowly twists the cap, with the rotation sound clearly audible; a small spoon scoops gel and drops it onto the glass with a soft 'plop' that reveals dense gel with tiny air bubbles; the hand presses onto the gel and spreads it in circular motions. Scene: a single light-glass surface against a dark background. Light: dramatic cool backlight making the gel glow like a gemstone. Camera: locked macro overhead and oblique, no shake. Style: photoreal product macro ASMR. Audio: pure crisp triggers, no music. Constraints: no watermark, no on-screen text, no extra hands.", "tags": [ "asmr", "skincare", "macro", "audio-driven" ], "params": { "aspect": "9:16", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 护肤品 ASMR (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-055", "model": "seedance-2.0", "category": "asmr", "title": "手部 ASMR 第一视角(ByteDance Seed 官方 T2V)", "prompt": "Immersive first-person hand ASMR video. Close-up shot, under warm, soft lighting. A pair of slender hands gently triggers different objects in sequence: the light scratching of frosted glass, rubbing of plush fabric, gentle tapping on an acrylic board, light pinching of bubble wrap, and lightly running fingers across a wooden comb's teeth. Finger movements are slow and gentle. Pure, natural trigger sounds with no background music; the visual atmosphere is relaxing and healing. Constraints: no watermark, no on-screen text, no extra hands.", "tags": [ "asmr", "pov", "hands", "tactile", "official" ], "params": { "aspect": "9:16", "duration": "15s", "resolution": "1080p" }, "source": { "name": "ByteDance Seed - Seedance 2.0 official launch blog (T2V #5)", "url": "https://seed.bytedance.com/en/blog/official-launch-of-seedance-2-0", "tier": "official" }, "notes": "ByteDance Seed 官方 T2V #5 (Hand ASMR),原文一字不改" }, { "id": "sd-056", "model": "seedance-2.0", "category": "asmr", "title": "微型烹饪剧场(黑背景)", "prompt": "Miniature cooking video on a pure black background with dramatic top lighting. Subject: full-sized hands working on miniature props. Action: hands carefully place a tiny cutting board and a fingernail-sized knife; slice a tiny garlic clove precisely; add a micro drop of olive oil to a mini frying pan over a tea candle; crack a small quail egg into the pan; nudge the egg with a tiny spatula; plate a miniature sunny-side-up egg. Scene: a flat dark surface with miniature kitchen props. Light: hard top key, deep black surround. Camera: locked overhead macro. Style: photoreal miniature cinematic. Audio: elegant classical music in the background; crisp micro-sounds (knife click, sizzle) synced to the music. Constraints: no watermark, no on-screen text, no extra hands.", "tags": [ "miniature", "cooking", "asmr", "classical" ], "params": { "aspect": "9:16", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 微型烹饪剧场 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-057", "model": "seedance-2.0", "category": "asmr", "title": "和牛牛排煎制特写", "prompt": "Subject: a thick-marbled wagyu steak and a hot cast-iron skillet. Action: the steak hits the pan with a heavy sizzle; visible fat renders along the marbling lines; tiny bubbles of oil pop on the surface; subtle smoke curls upward. Scene: a single dark stovetop, no distractions. Light: warm key from camera left, hard rim from the heat, deep dark background. Camera: locked extreme close-up at near-flush with the pan surface, shallow depth of field. Style: photoreal food macro ASMR. Audio: loud aggressive sizzling, faint timer ticking in the distance, no music. Constraints: no watermark, no on-screen text, no extra hands.", "tags": [ "food", "asmr", "macro", "sizzle" ], "params": { "aspect": "1:1", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 和牛特写 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-058", "model": "seedance-2.0", "category": "comedy", "title": "复联反高潮喜剧", "prompt": "Avengers Endgame big fight scene, cinematic style. Subject: Thanos plus several superheroes plus a stand-in Spider-Man character. Action: Thanos suddenly freezes the battle and holds up his hands looking genuinely apologetic; the superheroes pause looking surprised and slowly start to nod and walk away accepting his apology; then the Spider-Man stand-in shouts an off-screen line like 'Wait - he did kill a lot of people!'; everyone immediately turns back in exaggerated heroic poses and rushes at Thanos, comically tackling and kicking him while he is flustered and off-balance. Light: stormy sky with hard ratio and harsh accent lights. Camera: dynamic mid-shots and reaction close-ups with quick cuts on the punchline. Style: cinematic blockbuster parody, photoreal. Audio: cinematic orchestral cue then comedy sting on the turn. Constraints: avoid copyrighted character likenesses by 1:1 imitation - inspired-by only; no watermark, no Logo, no on-screen text.", "tags": [ "comedy", "parody", "dialog", "multi-character" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 复联反高潮喜剧 (Imagine.art 原稿 + 8 要素整理 + 添加版权安全约束)" }, { "id": "sd-059", "model": "seedance-2.0", "category": "comedy", "title": "水獭飞行员纪录片喜剧", "prompt": "Subject: a river otter wearing aviator goggles and a tiny scarf, paws on the controls of a small single-engine airplane. Action: the otter looks focused yet playful, swiveling its head to check instruments; the plane soars over rivers, forests, and mountains; the propeller spins with gentle motion. Scene: a sunny afternoon sky with patchy clouds. Light: bright golden hour key, soft cloud bounce fill. Camera: a dynamic low-angle tracking shot following the plane from outside, intercut with a tight cockpit close-up of the otter. Style: photoreal nature-doc-with-whimsy, cinematic framing. Audio: prop hum, gentle wind, light orchestral cue. Constraints: maintain otter identity across cuts, no watermark, no text, no Logo on the plane.", "tags": [ "comedy", "animal", "whimsical", "documentary" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 水獭飞行员 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-060", "model": "seedance-2.0", "category": "craft", "title": "老人制陶(双手特写纪录片)", "prompt": "Subject: an elderly craftsman's hands on a potter's wheel. Action: extreme close-ups alternate - fingers pressing into wet clay, water spreading across the spinning surface, the form rising and narrowing under steady pressure; the camera never cuts to his face, only hands. Scene: a small workshop with a window. Light: natural window key from one side, deep ambient shadow. Camera: locked macro and over-the-shoulder hand-level close-ups, no shake. Style: photoreal documentary, intimate. Audio: low hum of the wheel, wet clay sounds, distant birdsong from an open window, no music, no narration. Constraints: no watermark, no on-screen text, no extra hands.", "tags": [ "craft", "pottery", "hands", "documentary" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 老人制陶 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-061", "model": "seedance-2.0", "category": "craft", "title": "芭蕾排练后台纪录", "prompt": "Subject: a female principal ballet dancer in a studio. Action: she marks through a solo at half speed, then runs it full out; close-up of her feet in pointe shoes on the rosin-scuffed floor; a wide shot as she stops, breathes, and speaks quietly to the rehearsal director off-camera; the camera holds on her face as she prepares for another run - focused, expressionless, internal. Scene: a ballet rehearsal studio with mirrors lining one wall. Light: natural studio daylight from a high window. Camera: a slow handheld push as she crosses the studio; then locked-off for the close-up beats. Style: photoreal documentary cinematic. Audio: pointe shoes on Marley floor, controlled breathing, no music. Constraints: maintain face stable, no watermark, no on-screen text.", "tags": [ "dance", "ballet", "documentary", "rehearsal" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 芭蕾排练 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-062", "model": "seedance-2.0", "category": "craft", "title": "霓虹书法(魔法现实)", "prompt": "Subject: a female calligraphy artist at a small lit table in a dark studio. Action: she dips a large brush in ink and begins writing a single Chinese character on white paper; the camera pushes slowly in; as the brush moves, the ink trails begin to glow deep red and gold; the strokes lift fractionally off the paper, hovering with a faint luminescence; when the final stroke completes, all the strokes glow at once, then settle back to ordinary ink; she lifts the brush, pauses, and examines the character. Scene: a single pool of light on a clean wooden table, deep black studio around it. Light: a single tungsten practical above the table, deep dark background. Camera: a slow continuous push-in. Style: photoreal magic realism. Audio: only the sound of the brush on paper, no music. Constraints: keep the character readable; no watermark, no extra English text.", "tags": [ "craft", "calligraphy", "neon", "magic-realism" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Imagine.art - Seedance 2.0 prompt guide", "url": "https://www.imagine.art/blogs/seedance-2-0-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "保留 reputable - 霓虹书法 (Imagine.art 原稿 + 8 要素整理)" }, { "id": "sd-063", "model": "seedance-2.0", "category": "craft", "title": "@图1+@图2+@图3+@图4 治愈短片(ByteDance Seed 官方 R2V)", "prompt": "Refer to the shooting script in @Image1, and draw on the storyboard, shot scale, camera movement, visuals and copy in @Image1. The character is from @Image2, the scene is from @Image3, and the props are from @Image4. Create a 15-second healing short film. Constraints: maintain character identity from @Image2, environment palette from @Image3, props from @Image4; no watermark, no extra on-screen text, no warped subjects.", "tags": [ "r2v", "reference", "healing", "official", "multi-image" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "ByteDance Seed - Seedance 2.0 official launch blog (R2V #1)", "url": "https://seed.bytedance.com/en/blog/official-launch-of-seedance-2-0", "tier": "official" }, "notes": "ByteDance Seed 官方 R2V #1 - healing short film (4-image multi-reference),原文一字不改" }, { "id": "sd-064", "model": "seedance-2.0", "category": "craft", "title": "马上递花(ByteDance Seed 官方 R2V 视频延长)", "prompt": "Extend the video length. The camera follows a man in orange riding a brown horse. He speeds up and gallops to a large tree with orange flowers ahead, breaks off two blossoms from the branch. Then others ride into the frame one after another. The camera pushes in to shoot the man in orange clothes dismounting from the horse, then the camera quickly circles around him. He turns to walk toward a woman in white riding a white horse and presents the flowers to her. Style: Chinese classical lady painting, 3D, cheerful folk music, shadow puppet style, with black, white, and orange as the main color palette. Constraints: maintain the man-in-orange and woman-in-white identities across the extension, no watermark, no on-screen text.", "tags": [ "r2v", "video-extension", "chinese-classical", "official" ], "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "ByteDance Seed - Seedance 2.0 official launch blog (R2V #2)", "url": "https://seed.bytedance.com/en/blog/official-launch-of-seedance-2-0", "tier": "official" }, "notes": "ByteDance Seed 官方 R2V #2 - video extension(马上递花),原文一字不改" }, { "id": "sd-065", "model": "seedance-2.0", "title": "Burger Girl 变形(Format 1 三幕式)", "prompt": "Montage, multi-shot action Hollywood movie, don't use one camera angle or single cut, cinematic lighting, photorealistic, 35mm film quality, ARRI ALEXA aesthetic. A pink-haired girl with glasses, cream top and jeans sits on the hood of a white pickup truck under a concrete overpass at dusk, casually eating a burger. A pale zombie with wet dark hair sprints toward her from the shadows. The girl calmly sets down the burger, her body erupts into a massive pale tusked creature with clawed hands, devours the zombie whole, then shrinks back to human form and picks up the burger. Handheld shake throughout, dark comedy pacing with horror undertones. Total: 15s / 6 shots / 16:9. No 3D, no cartoon, no VFX.", "tags": [ "transformation", "three-act", "horror-comedy", "viral", "monster" ], "category": "viral-transform", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Higgsfield · Format 1 Burger Transformation", "url": "https://higgsfield.ai/s/seedance-2-0-higgsfieldai-kQBLPo", "tier": "reputable-3rd-party" } }, { "id": "sd-066", "model": "seedance-2.0", "title": "Bus 变形(Robo-Girl 反差)", "prompt": "Montage, multi-shot action Hollywood movie, cinematic lighting, photorealistic, 35mm film quality, ARRI ALEXA aesthetic. Inside a suburban bus packed with passengers, a red-haired girl in white tank top, cat-eye sunglasses and headphones sits unbothered while outside a colossal dark-blue rocky creature with sharp shoulder horns rampages through the street. All passengers press against the left windows in terror. The girl calmly opens the right window, climbs onto the roof, transforms into a green-armored robo-girl, destroys the creature with one massive blast, then drops back inside as a human. Handheld shake throughout, escalation arc from calm to explosive to calm. Total: 15s / 6 shots / 16:9.", "tags": [ "transformation", "robot", "monster", "viral", "anime-style" ], "category": "viral-transform", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Higgsfield · Format 1 Bus Transformation", "url": "https://higgsfield.ai/s/seedance-2-0-higgsfieldai-kQBLPo", "tier": "reputable-3rd-party" } }, { "id": "sd-067", "model": "seedance-2.0", "title": "Electro POV — 第一视角电力幻想", "prompt": "Single continuous shot, first-person POV perspective, the camera IS her eyes, hyper-chaotic handheld motion, completely unstabilized, violent raw human movement, constant micro-jitters, aggressive head swings, abrupt jerks, frequent over-rotation, wide-angle lens (strong distortion), 15 seconds, her hands always visible in frame, no music only raw SFX, cinematic lighting, photorealistic, grounded realism, 35mm film look, heavy film grain, ARRI ALEXA aesthetic, practical VFX feel. In a storm-soaked industrial ruin, her hand snaps forward catching a crackling violet lightning sphere, electricity arcing violently, she crushes it instantly — blinding energy surges through her fingers as fractal lightning veins explode across both forearms. Enemies emerge — jagged obsidian creatures crawl across walls, while a colossal iron titan rises behind them. She lunges forward, slamming both hands down releasing a radial lightning surge, the titan retaliates with a beam, she catches it with one hand and redirects it upward while sprinting at the titan, drives both hands into its core, RAMPS TO SLOW MOTION as the core fractures, energy tearing outward in layered shockwaves. SFX: electric crackle, energy burst, deep metallic titan rise, slow-motion electric hum. Total: 15s / 1 shot / 16:9.", "tags": [ "pov", "first-person", "powers", "lightning", "epic", "action" ], "category": "fantasy-scifi", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Higgsfield · Format 2 Orbs Electro", "url": "https://higgsfield.ai/s/seedance-2-0-higgsfieldai-kQBLPo", "tier": "reputable-3rd-party" } }, { "id": "sd-068", "model": "seedance-2.0", "title": "Gladiator POV(罗马角斗士)", "prompt": "One continuous shot, POV gladiator perspective in the Colosseum arena, no cuts, no zoom, natural head movement, a furious enemy warrior sprints straight toward the camera through thick dust and sunlight, heavy footsteps shaking the ground, as he reaches close the POV character reacts instantly — grabs him mid-charge and slams him violently onto the sand, powerful impact, dust explosion, no pause, immediate chaos erupts all around, multiple gladiators fighting simultaneously, intense close combat everywhere, swords clashing, shields smashing, archers releasing arrows overhead, a chariot bursts through the scene, POV character fights barehanded with raw force — blocking strikes, grabbing opponents, throwing them aside, camera shaking from impacts, breath sounds, dust in the air, dramatic sunlight beams cutting through shadows, epic scale, photorealistic, motion blur on hits. Total: 15s / 1 shot / 16:9.", "tags": [ "pov", "gladiator", "epic", "fight", "single-shot" ], "category": "action", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Higgsfield · Format 3 POV Gladiator", "url": "https://higgsfield.ai/s/seedance-2-0-higgsfieldai-kQBLPo", "tier": "reputable-3rd-party" } }, { "id": "sd-069", "model": "seedance-2.0", "title": "中世纪骑士 POV(极简版)", "prompt": "A single-frame POV video of a medieval knight riding a horse with a sledgehammer in his hands, riding and fighting epically, smashing his opponents with a sledgehammer while riding a horse, to make it look realistic with blood.", "tags": [ "pov", "medieval", "knight", "minimal-prompt", "epic" ], "category": "action", "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Higgsfield · Format 3 Medieval Knight (Minimal)", "url": "https://higgsfield.ai/s/seedance-2-0-higgsfieldai-kQBLPo", "tier": "reputable-3rd-party" } }, { "id": "sd-070", "model": "seedance-2.0", "title": "火车顶武士对决(Train Rooftop Fight)", "prompt": "Single continuous shot 15s: The camera begins below the train roof level — an extreme low angle looking up, the grey sky above and the train's metal roof surface filling the upper frame, power line cables visible on both sides, electric sparks already arcing between the cables and the roof surface. The FPV arm accelerates upward and crests the rooftop edge just as both female samurai warriors sprint into frame from opposite ends — dark armor catching the electric flash light, katanas already drawn — the camera dropping to roof level and sweeping into a full 360-degree orbit around both fighters as they clash center-roof, blades ringing, feet sliding on the curved metal surface, electric sparks erupting in curtains. The orbit completes and camera settles tight behind the first warrior as she draws back her katana in a full overhead swing — motion ramping into deep slow motion, the second warrior already reading the strike, arching her entire body backward in a deep spine bend, the blade passing centimeters from her face, a single small lock of dark hair separating in the slow motion wind. Motion snaps back as the second warrior drives her elbow into the first warrior's chest, both fighters exchange rapid blade-locked strikes, then lock arms, pivot, and drive each other off the edge — both bodies leaving the roof, tumbling sideways, camera descending with them, both hitting the river surface in twin white explosions, camera plunges beneath the surface — the world going deep blue-green, sunlight filtering down in shafts, both warriors still moving through the underwater space, the fight continuing in the silent slow drift of the river current. NO MUSIC / NO SFX. Total: 15s / 1 shot / 16:9.", "tags": [ "fight", "samurai", "train", "slow-motion", "epic", "single-shot" ], "category": "action", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Higgsfield · Format 4 Train Rooftop", "url": "https://higgsfield.ai/s/seedance-2-0-higgsfieldai-kQBLPo", "tier": "reputable-3rd-party" } }, { "id": "sd-071", "model": "seedance-2.0", "title": "Lollipop Girl 工业厂房格斗", "prompt": "Medium wide shot in a dark industrial warehouse, muscular bald man with black tribal face paint and open camouflage tactical vest over bare chest — squares off against a petite young woman half his size with messy dark hair and bangs, oversized gray sweatshirt slipping off one shoulder, dark pleated mini skirt, knee-high black socks, lollipop stick hanging from her lips. The big man swings a massive haymaker — she ducks under it effortlessly and delivers a rapid three-punch combo to his gut, the painted fighter doubles over in shock, she springboards off a metal crate and leaps onto his back wrapping both arms around his neck in a chokehold, the big man spins wildly trying to shake her off, crashing into steel shelving units that collapse in a domino chain, he finally grabs her by the back of her sweatshirt and hurls her across the room — she flips mid-air and lands on her feet sliding backward, pulls the lollipop from her mouth, grins, charges back in low to the ground weaving between his legs and delivers a devastating uppercut, the painted fighter's eyes go wide, slow-motion knees buckling, she winds up a running flying right hook that connects with the jaw sending him toppling backward like a felled tree through a stack of wooden pallets, dust cloud erupting on impact, harsh single-source overhead industrial light, warm amber on skin with cool steel-blue environment, anamorphic 35mm, Guy Ritchie speed-ramping with Snyder impact slow-motion. Total: 15s / 1 shot / 16:9.", "tags": [ "fight", "underdog", "industrial", "speed-ramp", "anamorphic" ], "category": "action", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "Higgsfield · Format 4 Lollipop Girl", "url": "https://higgsfield.ai/s/seedance-2-0-higgsfieldai-kQBLPo", "tier": "reputable-3rd-party" } }, { "id": "sd-072", "model": "seedance-2.0", "title": "极简动画 prompt(Format 5)", "prompt": "Fight of a 3D person with 2D", "tags": [ "animation", "minimal-prompt", "ai-creative", "fight" ], "category": "creative", "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Higgsfield · Format 5 Animation", "url": "https://higgsfield.ai/s/seedance-2-0-higgsfieldai-kQBLPo", "tier": "reputable-3rd-party" } }, { "id": "sd-073", "model": "seedance-2.0", "title": "动作迁移(最简公式)", "prompt": "Use the motion from the video in @video_1 with the characters in @image_1", "tags": [ "motion-transfer", "minimal-prompt", "omni-reference", "v2v" ], "category": "creative", "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · 100+ Prompts", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-074", "model": "seedance-2.0", "title": "Eclipse 耳机产品广告(Macro 多镜)", "prompt": "TITLE: Eclipse Headphones. TONE: Premium minimalism, elegant tension. STYLE: Photoreal macro product cinematography. FORMAT: 16:9. DURATION: 8 seconds. [0–2s HOOK] Macro 85mm shot gliding over brushed-metal ear-cup. Soft edge light wraps around logo. Mist of particles drifts. Ambient hum + low bass swell. [2–4s SETUP] 35mm side tracking — model's silhouette lifts headphones in slow motion. Light gradient fades from blue → gold. Sub-bass pulse builds. [4–6s ACTION] Tight 50mm close-up as ear-cup seals over ear. Sound drops to silence. Camera pushes in microscopically, reflection catching logo flare. [6–8s RESOLUTION] Wide 24mm reveal — subject lost in soundscape; particles dissolve into brand emblem. Subtle logo sting + fade. Continuity: Soft diffused key light, desaturated palette with gold accents, macro focus transitions.", "tags": [ "product", "headphones", "macro", "luxury", "multi-shot" ], "category": "product-commercial", "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · 100+ Prompts (Eclipse Headphones)", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-075", "model": "seedance-2.0", "title": "魔杖亲子广告(UGC 风格)", "prompt": "0:00–0:03 Parent holds wand up to camera. 'This arrived yesterday.' Casual, no setup needed. 0:03–0:10 Cut to kid holding wand, pointing it at the TV. Completely serious. 'Lumos.' Clicks the light switch instead — light turns on. Kid loses their mind. Stage it so the timing works. Hold on the kid's face when the light comes on. 0:10–0:13 Parent back on camera, whispering: 'They think they're actually magic. I'm not correcting them.' Conspiratorial. Warm. 0:13–0:15 Wand close-up. Text overlay: 'Link in bio.' Hard cut.", "tags": [ "product", "ugc", "family", "wand", "social-viral" ], "category": "product-commercial", "params": { "aspect": "9:16", "duration": "15s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · 100+ Prompts (Wand UGC)", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-076", "model": "seedance-2.0", "title": "铁匠锻造戒指(叙事独白)", "prompt": "A handheld, front-facing shot inside a dark medieval forge lit by roaring fire and molten metal glow. Warm orange light flickers across stone walls, hanging blacksmith tools, and drifting sparks. Mild handheld shake, authentic candid phone-camera vibe. A rugged blacksmith in a worn leather apron stands beside the furnace, soot smudged across his face. He lifts a freshly forged golden ring toward the camera. The metal glows faintly from the heat. Fine engraved markings shimmer. Blacksmith (natural, slightly amazed): 'Alright… wait until you see this.' He tilts the ring closer, slowly rotating. Blacksmith (curious): 'This is the one ring. Just came out of the forge.' He turns it again, the molten glow reflecting in his eyes. Blacksmith (surprised, playful): 'Yeah… that turned out far better than I expected.' He lowers the ring slightly, leaning toward the camera. Blacksmith (casual, friendly): 'The gold's perfect, the finish is clean, and these markings… look at that. Honestly, this feels like something out of legend.' He raises the ring again so it glows dramatically. Blacksmith (smiling, confident): 'The one ring. Finest piece I've ever made.' The shot ends naturally.", "tags": [ "product", "narrative", "blacksmith", "monologue", "medieval" ], "category": "product-commercial", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · 100+ Prompts (Blacksmith Ring)", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-077", "model": "seedance-2.0", "title": "Snap Shockwave 时间冻结都市", "prompt": "Use <<>> as the main character, keeping facial features and body proportions consistent throughout. Cinematic time-freeze short film, 15 seconds, ultra-realistic, shot on Arri Alexa Mini, 50mm lens, natural daylight with hard shadows, shallow depth of field. [0:00–0:03] A busy city sidewalk, time flowing normally. Steadicam frontal medium shot tracking the main character walking confidently through the crowd. As he walks, he raises his right hand and snaps his fingers. [0:03–0:06] At the exact moment of the snap — a powerful white spherical shockwave bursts from his fingertips, with visible air distortion and light refraction, rapidly expanding outward. It sweeps across the ground, lifting dust, instantly freezing them mid-air. It hits pedestrians — each freezes mid-step. It reaches the sky — every pigeon freezes mid-flight. After the wave dissipates — absolute silence. Tiny time-frozen particles float in the air. [0:06–0:09] In the silence, only the character's footsteps echo crisply. The camera moves backward, flying in front of him as he walks toward it, casually passing through the frozen crowd. He notices a pigeon suspended above him, reaches up, and gently touches its feathers. The pigeon remains completely still but emits a low cooing sound. He continues walking. [0:09–0:15] He turns, pulls a cigarette from his jacket pocket, places it in his mouth, flicks a lighter — click — lights it, takes a drag, slowly exhales smoke. The smoke is the only thing moving in the frozen world. He turns to face the camera with confident smile, raises his right hand, and snaps again — everyone explodes back into motion.", "tags": [ "bullet-time", "snap-shockwave", "powers", "urban", "cinematic" ], "category": "fantasy-scifi", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · 100+ Prompts (Snap Shockwave)", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-078", "model": "seedance-2.0", "title": "360° 年龄循环(童年到老年)", "prompt": "Style: Cinematic realism. Orbit Rule: Slow, steady clockwise 360°. Length: ~15 seconds. CAMERA: Camera begins orbiting clockwise around a single character standing in open field at golden hour. Orbit speed constant. No cuts. Character ages progressively across full rotation. [0–3s] Child (Age 8): Small child standing in tall grass holding a paper airplane. Soft warm sunlight. Child: 'I'll go further than anyone.' Hair subtly grows longer, posture shifts. [3–6s] Teen (Age 16): Same field, sunset slightly deeper. Teen now holds school backpack. Jawline sharper. Teen: 'I won't quit.' Camera keeps orbiting. Clothes shift mid-rotation. [6–9s] Young Adult (Age 25): Field now dusk. Slight wind. Adult in casual jacket. Adult: 'I hope this is right.' Orbit continues — grey streak appears in hair. [9–12s] Middle Age (45): Night settling in. Wrinkles form. Middle-aged: 'I did what I could.' Orbit continues — skin texture thins. [12–15s] Elderly (75+): Full night sky overhead. Older figure, slower breathing, slight tremble. He looks upward. Elderly (whisper): 'And it was enough.' Orbit completes facing him under stars. Ends with a space logo with the text 'PROJECT HAIL GARY'. Fade out.", "tags": [ "orbit", "time-passage", "aging", "emotional", "single-shot" ], "category": "cinematic", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · 100+ Prompts (Age Loop)", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-079", "model": "seedance-2.0", "title": "Transformer 飞机变机器鹰", "prompt": "An airplane is parked in a dilapidated parking lot in outskirt of Jakarta. The airplane begins to violently disassemble and reorganize from the inside out, transforming into a grumpy, evil Cybertronian mechanical giant eagle. 00–02s The Setup: Wide shot, eye-level. An airplane slowly drives forward on a dusty lot under a clear, sunny sky. The tone is grounded and realistic. 02–05s The Transformation Begins: Dynamic VFX action shot. The airplane violently bursts open. The front splits as a massive metallic silver eagle head with glowing yellow eyes pushes through. The vehicle's chassis, doors rapidly fold, spin, and reconfigure into heavy mechanical limbs. 05–08s The Ejection: Fast-paced tracking shot. The driver is thrown out of the transforming tank. He tumbles onto the dusty ground as the machine finishes shifting into a giant robotic eagle, kicking up dirt. 08–11s The Mount: Action shot. The driver scrambles to his feet, charges at the colossal beast, grabs onto the shifting side panels, vaults himself high into the air, landing squarely on the eagle's metallic back. 11–14s The Hero Pose: Low angle, epic hero shot. The man straddles the giant robotic eagle, gripping the armor tightly. The mechanical eagle stands aggressively, jaws open in a fierce roar. A cinematic lens flare hits the lens.", "tags": [ "transformation", "transformer", "vfx", "cybertronian", "epic" ], "category": "fantasy-scifi", "params": { "aspect": "16:9", "duration": "14s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · 100+ Prompts (Transformer)", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-080", "model": "seedance-2.0", "title": "VFX 金属手臂磁吸(V2V 极简)", "prompt": "after 0.2 seconds metal armour magnetically attaches itself to the arm and hand in @video1", "tags": [ "vfx", "v2v", "minimal-prompt", "armor", "magnetic" ], "category": "creative", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · VFX Series", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-081", "model": "seedance-2.0", "title": "VFX 黑色 Goo 触手覆盖手臂", "prompt": "the arm and hand in @video_1 start to get covered in shiny black goo, black tendrils crawl up the arm", "tags": [ "vfx", "v2v", "minimal-prompt", "symbiote", "horror" ], "category": "creative", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · VFX Series", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-082", "model": "seedance-2.0", "title": "VFX 透明手臂(显示骨骼肌肉)", "prompt": "the arm and hand in @video_1 start to go translucent revealing the bones and muscle in the arm and hand", "tags": [ "vfx", "v2v", "minimal-prompt", "x-ray", "anatomy" ], "category": "creative", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · VFX Series", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-083", "model": "seedance-2.0", "title": "VFX 路面塌陷怪物涌出", "prompt": "a huge sinkhole is in the road in @video1, monsters start crawling out from the sinkhole, monster creature sounds and audio", "tags": [ "vfx", "v2v", "minimal-prompt", "monster", "horror" ], "category": "creative", "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · VFX Series", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-084", "model": "seedance-2.0", "title": "FPV 飞行 — 曼哈顿穿越写字楼(Start+End Frame)", "prompt": "Use Image 1 as the exact starting frame and Image 2 as the exact final frame. Do not change either image. The video must start matching Image 1 and end matching Image 2 perfectly. Generate a continuous FPV camera shot lasting 12 seconds. The camera represents a first-person viewpoint. Do not show any drone, propellers, shadows, or physical camera rigs. The camera moves fast but remains controlled and cinematic. Use smooth forward motion with gentle curves and subtle banking. Keep the horizon stable and level. No spinning, no flips, no full rotations, no aggressive rolling. Start at the Manhattan skyline from Image 1. Move forward and descend toward a nearby office building in one smooth motion. Fly into the building through the windows and pass through interior office spaces, moving past tables, chairs, monitors, and objects with realistic scale. Exit the building and continue descending along the exterior toward the main road. At street level, move forward past traffic including yellow taxis, cars, and pedestrians. Continue forward toward a dense commercial area resembling Times Square. Slightly rise toward detailed billboards and screens, then perform a controlled directional turn around them. After the turn, descend back toward street level and move toward a dentist that matches the final scene. Fly through the entrance with a smooth forward glide. Inside, pass by tables, people, and interior elements with realistic lighting. Gradually slow down only at the very end and settle perfectly into the exact framing of Image 2.", "tags": [ "fpv", "start-end-frame", "manhattan", "long-shot", "cinematic" ], "category": "cinematic", "params": { "aspect": "16:9", "duration": "12s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · 100+ Prompts (FPV Manhattan)", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-085", "model": "seedance-2.0", "title": "JSON 格式:龙的第一视角空战", "prompt": "{\"title\": \"Cinematic First-Person Dragon Battle Flight\", \"style\": \"epic fantasy cinematic, ultra-realistic\", \"camera\": {\"perspective\": \"first-person POV of a dragon\", \"shot_type\": \"continuous aerial combat shot\", \"lens\": \"wide anamorphic fantasy lens\", \"focal_length\": \"14mm\", \"motion\": [\"massive swooping dives\", \"violent barrel rolls\", \"powerful wing beats\"]}, \"scene\": {\"environment\": \"medieval battlefield at sunset\", \"details\": [\"burning siege towers\", \"soldiers below\", \"catapults firing\", \"smoke columns\"]}, \"action\": {\"movement\": [\"dragon dives toward battlefield\", \"flies low over armies\", \"ascends sharply above castle walls\"], \"events\": [{\"type\": \"fire_breath\", \"description\": \"dragon unleashes massive wall of fire\"}, {\"type\": \"arrow_storm\", \"description\": \"thousands of flaming arrows rush toward camera\"}]}}", "tags": [ "json-format", "dragon", "pov", "aerial-combat", "structured" ], "category": "fantasy-scifi", "params": { "aspect": "16:9", "duration": "12s", "resolution": "1080p" }, "source": { "name": "YouTube PypS8xF5lfo · 100+ Prompts (JSON Dragon)", "url": "https://docs.google.com/document/d/1DyulbGoZMY0dfXQrrph0KThf9XzmGU7Ceo7HFdyFbns/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-086", "model": "seedance-2.0", "title": "微型女孩极速穿越玩具世界", "prompt": "On the floor of a children's room, a miniature-sized girl rides a tiny skateboard, zipping along at high speed just above the ground. The sense of scale makes human-sized toys and furniture appear gigantic. The camera tracks from a low angle right behind the protagonist, creating a continuous one-shot style video that constantly moves forward. Ultra-wide-angle lens, motion blur, depth of field, cinematic lighting. Stage 1: She dashes at high speed through a narrow passage resembling a canyon of buildings in a Lego city, slaloming forward between building blocks. Stage 2: A giant ball rolls toward her from the front and she slides into a tiny gap against the wall, ducks under falling building blocks by a hair's breadth, and barely passes between the tires of a moving toy car. Stage 3: She bursts through the gap between the pages of a picture book blowing shut in the wind, slips under the arm of a swaying stuffed animal, and finally dives into the slight opening of a closing toy box lid, disappearing into the dark depths at maximum speed. A thrilling and exhilarating video featuring a continuous series of narrow escapes. An immersive, theme-park-ride-like experience.", "tags": [ "miniature", "scale", "chase", "theme-park", "continuous" ], "category": "fantasy-scifi", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "YouTube DqWmHvIIcww · 50 Viral Prompts", "url": "https://docs.google.com/document/d/1Q9-oa6zlZeg4eq5xn_axpApuWdcczj60GDG_Fpd27Bk/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-087", "model": "seedance-2.0", "title": "二战诺曼底登陆 POV", "prompt": "15-second continuous single-shot action sequence. No cuts. No scene transitions. Cinematic war realism, handheld POV camera, desaturated colors, heavy grain, harsh daylight through smoke, water droplets on lens, mud and debris, realistic physics, chaotic but readable motion. Extreme immersion, grounded soldier movement, controlled disorientation, no stylized action. Scene: WWII beach landing, shallow water, incoming waves, heavy smoke, distant bunkers firing, explosions across shoreline. 0–3s arrival: POV from a soldier jumping into shallow water. Camera drops abruptly: water splashes across lens, breathing heavy, vision unstable. Immediate gunfire hitting water nearby. 3–6s disorientation: Camera rises unevenly. Running forward through water. Bullets impacting around, soldiers falling in peripheral vision, explosions sending sand and water upward. 6–10s push forward: Transition from water to wet sand. Soldier stumbling, regaining balance. Nearby cover. Quick crouch. Hands briefly visible steadying weapon. 10–12s slow motion shock: Explosion nearby. Time slows: sand and debris suspended, shockwave rippling, sound implied muted. Camera slightly tilts. 12–15s survival instinct: Time snaps back. Soldier rises and pushes forward again toward cover. Gunfire continues. Camera lurches forward with urgency. Freeze.", "tags": [ "pov", "wwii", "normandy", "war", "single-shot", "speed-ramp" ], "category": "cinematic", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "YouTube DqWmHvIIcww · 50 Viral Prompts", "url": "https://docs.google.com/document/d/1Q9-oa6zlZeg4eq5xn_axpApuWdcczj60GDG_Fpd27Bk/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-088", "model": "seedance-2.0", "title": "武侠竹林剑斗(中文风格)", "prompt": "Style: Chinese wuxia martial arts, expressive and sharp ink-style aesthetics. Lighting: cinematic lighting, cold moonlight with soft diffuse shadows. 0–3s: A white-robed swordsman (appearance reference: Image 1) faces a black-clad assassin (appearance reference: Image 2) in a moonlit bamboo forest. A half crescent moon hangs in the sky, casting cool-toned diffuse moonlight. Wide cinematic shot. Bamboo leaves slowly fall through the air. 4–7s: The black-clad assassin suddenly dashes forward, moving with explosive speed. His short blade slices three intersecting golden arcs through the air. The white-robed swordsman leans back and slides, lightly touching his thin sword to the ground to pivot and spin away. Medium tracking shot, fast and fluid movement. 8–12s: The fighters engage in a high-speed sword exchange. Blades collide with bursts of sparks, metal striking metal. The camera orbits around the duel, capturing the intensity from multiple angles. Motion afterimages trail their movements, sword energy sweeps through the bamboo forest. 13–15s: The white-robed swordsman leaps into the air, unleashing a powerful white sword energy slash. The black-clad assassin blocks and staggers backward. The shot freezes momentarily with a subtle camera shake. Visual Style: Tsui Hark-style wuxia cinematic action, 4K ultra-HD, 60 FPS smooth motion.", "tags": [ "wuxia", "chinese", "swords", "bamboo-forest", "tsui-hark" ], "category": "action", "params": { "aspect": "16:9", "duration": "15s", "resolution": "4K" }, "source": { "name": "YouTube DqWmHvIIcww · 50 Viral Prompts (Wuxia)", "url": "https://docs.google.com/document/d/1Q9-oa6zlZeg4eq5xn_axpApuWdcczj60GDG_Fpd27Bk/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-089", "model": "seedance-2.0", "title": "东南亚夜市不可能运镜(ground-to-aerial)", "prompt": "FORMAT: 15s / single continuous impossible camera move / no dialogue. STYLE: Dense Southeast Asian night market, wet stone, steam and fire, orange lantern light, photorealistic ground-to-aerial cinematic 8K. Shot 01 (0:00–2:00): Camera starts at ankle level. Forest of feet on glistening wet stone. Market sounds overwhelming. Camera weaves between legs like water. Audio: Wet footsteps, distant vendor calls. Shot 02 (2:00–3:30): Camera rises slowly past steaming woks. A gas flame bursts at exact eye level — camera briefly engulfed in orange fire then emerges through it, uncut. Shot 03 (3:30–5:00): Camera weaves through hanging lanterns at mid-height. Red and orange light strobing. Shot 04 (5:00–6:30): Camera dips suddenly under a low table. A child has fallen asleep on a bag here. Quiet pocket of stillness. Shot 05 (6:30–8:00): Camera rises back up through the smoke of a charcoal grill — emerges above stall canopy. Shot 06 (8:00–10:00): Camera continues rising — now above rooftop level. The market reveals itself as an endless orange lantern sea stretching to the horizon. Shot 07 (10:00–12:00): Camera begins descending back down — targeting a single stall at the market edge. Shot 08 (12:00–13:30): Camera settles at the vendor's hands. Close on coins being stacked methodically. Shot 09 (13:30–15:00): Camera tilts up slowly to the vendor's face. They look up — directly into the camera. Hold. Fade to black. STYLE NOTES: Ground level claustrophobic. Aerial shot is the emotional exhale. Warm orange and red throughout. 8K.", "tags": [ "impossible-camera", "night-market", "ground-to-aerial", "asia", "single-shot" ], "category": "cinematic", "params": { "aspect": "16:9", "duration": "15s", "resolution": "4K" }, "source": { "name": "YouTube DqWmHvIIcww · 50 Viral Prompts (Night Market)", "url": "https://docs.google.com/document/d/1Q9-oa6zlZeg4eq5xn_axpApuWdcczj60GDG_Fpd27Bk/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-090", "model": "seedance-2.0", "title": "一战战壕不可能运镜", "prompt": "FORMAT: 15s / single continuous impossible camera move / no dialogue. STYLE: World War I trench warfare, deep mud, pouring rain, splintered wood, photorealistic ground-to-aerial cinematic 8K. Shot 01 (0:00–2:00): Camera starts at mud level. A forest of heavy mud-caked boots marching and sinking into sludge. Water filling footprints. Audio: Squelching mud, deafening artillery thuds. Shot 02 (2:00–3:30): Camera rises slowly past a sandbag wall. An artillery shell explodes just over the parapet — camera briefly engulfed in flying dirt and flash-fire, emerging uncut. Audio: Earth-shattering BOOM. Ringing ears. Shot 04 (5:00–6:30): Camera dips suddenly under a shattered wooden duckboard. A single pristine white field mouse curled up, shivering. Audio: Tiny heartbeats. Shot 06 (8:00–10:00): Camera continues rising — high above No Man's Land. The battlefield reveals itself as an endless, scarred lunar landscape of zigzagging trenches stretching to a blackened horizon. Audio: War sounds fading to mournful wind. Shot 09 (13:30–15:00): Camera tilts up slowly to the medic's hollow eyes. They look up — directly into the camera. Hold. Fade to black. Audio: A single drop of water hitting a tin cup. Silence. STYLE NOTES: Desaturated browns, grays, toxic yellows. 8K.", "tags": [ "impossible-camera", "wwi", "trench-warfare", "ground-to-aerial", "single-shot" ], "category": "cinematic", "params": { "aspect": "16:9", "duration": "15s", "resolution": "4K" }, "source": { "name": "YouTube DqWmHvIIcww · 50 Viral Prompts (WWI Trench)", "url": "https://docs.google.com/document/d/1Q9-oa6zlZeg4eq5xn_axpApuWdcczj60GDG_Fpd27Bk/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-091", "model": "seedance-2.0", "title": "女宇航员太空站逃生", "prompt": "A female astronaut in a white EVA suit sprints across the outer ring of a damaged orbital station above Earth. The station slowly rotates while massive solar panels break apart and drift into space. At the 2-second mark she jumps from the rotating ring toward a drifting cargo module. Camera floats beside her as Earth spins in the background. She grabs a handle, swings through the airlock hatch, and seals the door just as debris slams into the station behind her. Orbital debris field, zero-gravity leap, rotating space station exterior run, cinematic space lighting, 4K.", "tags": [ "space", "astronaut", "action", "zero-gravity", "escape" ], "category": "fantasy-scifi", "params": { "aspect": "16:9", "duration": "10s", "resolution": "4K" }, "source": { "name": "YouTube DqWmHvIIcww · 50 Viral Prompts (Astronaut)", "url": "https://docs.google.com/document/d/1Q9-oa6zlZeg4eq5xn_axpApuWdcczj60GDG_Fpd27Bk/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-092", "model": "seedance-2.0", "title": "豪华邮轮穿越不可能运镜", "prompt": "Continuous single take, camera tilting, rolling, spinning, flying through a colossal luxury cruise ship. Starts flying over the sunny upper deck dodging beach balls and sunbathers, camera dives straight down a twisting, translucent water slide, gliding past screaming children before splashing into the main pool. Sweeps underwater, effortlessly passing through a glass floor directly into the elegant grand dining room below, weaving through crystal chandeliers, clinking champagne glasses, and waiters balancing silver trays. Camera dives past the spinning kitchen doors into the humming, immaculate engine control room, transitions out the stern exhaust vent, gliding over the churning, foamy white wake as the sun sets over the ocean. No cuts, impossible camera moves, seamless transitions, energetic, dynamic, cinematic. 8K high definition, high quality footage.", "tags": [ "impossible-camera", "cruise-ship", "luxury", "single-shot", "vacation" ], "category": "cinematic", "params": { "aspect": "16:9", "duration": "15s", "resolution": "4K" }, "source": { "name": "YouTube DqWmHvIIcww · 50 Viral Prompts (Cruise)", "url": "https://docs.google.com/document/d/1Q9-oa6zlZeg4eq5xn_axpApuWdcczj60GDG_Fpd27Bk/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-093", "model": "seedance-2.0", "title": "亚马逊雨林穿越不可能运镜", "prompt": "Continuous single take, camera tilting, rolling, spinning, flying through a dense Amazonian rainforest ecosystem. Starts flying through the misty canopy dodging hanging vines and a sleeping sloth, camera dives downward into the hollow trunk of a massive ancient tree, gliding past bioluminescent fungi and a colony of bats. Sweeps out onto the humid forest floor, closely following a rapid line of leaf-cutter ants carrying green fragments, weaving through giant ferns and dew drops. Camera dives directly into a fast-flowing jungle river, transitioning smoothly underwater past darting piranhas and swaying aquatic plants, before bursting out of a roaring waterfall into a breathtaking aerial view of the infinite green jungle at dawn. No cuts, impossible camera moves, seamless transitions, energetic, dynamic, cinematic. 8K high definition, high quality footage.", "tags": [ "impossible-camera", "rainforest", "amazon", "nature", "single-shot" ], "category": "nature", "params": { "aspect": "16:9", "duration": "15s", "resolution": "4K" }, "source": { "name": "YouTube DqWmHvIIcww · 50 Viral Prompts (Amazon)", "url": "https://docs.google.com/document/d/1Q9-oa6zlZeg4eq5xn_axpApuWdcczj60GDG_Fpd27Bk/edit", "tier": "reputable-3rd-party" } }, { "id": "sd-094", "model": "seedance-2.0", "title": "Nano Banana 角色表降真实度(过审 prompt)", "prompt": "Convert this character sheet into a slightly stylized game character render. Keep the face highly recognizable, preserve facial features, bone structure, realistic proportions, and subtle skin detail. Reduce photorealism slightly. Use cleaner shapes, polished game-art rendering, and slightly simplified textures. Keep the original composition, pose, outfit, and character design unchanged. High-quality modern AAA video game character art.", "tags": [ "nano-banana", "bypass-face", "character-sheet", "preprocessing", "workflow" ], "category": "creative", "params": { "aspect": "1:1", "duration": "n/a (image preprocessing)", "resolution": "n/a" }, "source": { "name": "YouTube DqWmHvIIcww · Bypass Face Filter", "url": "https://www.youtube.com/watch?v=DqWmHvIIcww", "tier": "reputable-3rd-party" } }, { "id": "sd-095", "model": "seedance-2.0", "category": "product-commercial", "title": "产品开箱(Dreamina 官方品牌营销范式)", "prompt": "A premium [product] unboxing sequence. Subject: a sleek [product] resting inside pristine white packaging with embossed logo. Action: hands enter from the top, gently lift the inner tray and pull the device free in slow, deliberate motion. Scene: an infinite white cyclorama, subtle floor shadow under the box. Light: large overhead softbox key with a low-power back rim catching the bezel edge. Camera: locked-off three-quarter close-up, then a gentle 5-second push-in toward the product surface, no shake. Style: high-end product CF, photorealistic materials, brand colors visible on the box. Constraints: no watermark, no text, no Logo other than the brand, no extra hands, no warped packaging.", "tags": [ "product", "unboxing", "studio", "official", "8-element" ], "params": { "aspect": "9:16", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Dreamina (ByteDance / CapCut) - Seedance 2.0 prompt guide", "url": "https://dreamina.capcut.com/resource/seedance-2-0-prompt", "tier": "official" }, "notes": "改编自 Dreamina 官方品牌营销范式 #10 - product unboxing close-up + brand logo panning" }, { "id": "sd-096", "model": "seedance-2.0", "title": "多 keyframe 漫游(Text+Reference 替代 Start/End Frame)", "prompt": "Character walks from image4, which is where this whole video starts, to image2 (slightly right of the main studio area), then over to image5 (further right), and then finally to image1 (fully around to the back, pointing to a room full of monitors). While he walks seamlessly from scene to scene, he is saying \"Let me show you around. This is the main studio, and over here we have studio B, and next is our famous wall of monitors. Now, if you'll excuse me, I'm tired of you.\" Then he walks away.", "tags": [ "multi-keyframe", "reference-images", "walking-tour", "lip-sync", "studio-environment" ], "category": "creative", "params": { "aspect": "16:9", "duration": "10s+", "resolution": "1080p", "references": "4 images (image1..image5)" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-097", "model": "seedance-2.0", "title": "图片局部元素精确引用(@局部衣物/道具)", "prompt": "the man in image1 wearing the overalls from image3, carrying the shirt from image4", "tags": [ "reference-images", "local-attribute", "wardrobe-mix", "snippet" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "references": "3 images" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-098", "model": "seedance-2.0", "title": "图片当成视觉风格引用(stop-motion/插画风继承)", "prompt": "using image1 as a reference for the visual style", "tags": [ "style-transfer", "reference-images", "visual-style", "snippet" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "references": "1 style image" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-099", "model": "seedance-2.0", "title": "Lip sync 加文字稿双重约束(避免听错词)", "prompt": "The man in image1 is lip syncing to audio1; he says: \"Welcome back to the channel...\"", "tags": [ "lip-sync", "audio-reference", "transcript", "dual-anchor" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "references": "1 image + 1 audio" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-100", "model": "seedance-2.0", "title": "Multi-person Lip Sync(双角色对话自动分声线)", "prompt": "the dialogue between the two characters is lip sync driven by audio1", "tags": [ "lip-sync", "multi-person", "audio-reference", "snippet" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "references": "1 audio (with 2 voices)" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-101", "model": "seedance-2.0", "title": "音频作为背景音乐(AI 自动控制音量让对话可听)", "prompt": "using audio1 as background music", "tags": [ "background-music", "audio-reference", "auto-mixing", "snippet" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "references": "1 audio" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-102", "model": "seedance-2.0", "title": "声音克隆 vocal 一致性(不同场景同一嗓音)", "prompt": "use the vocal qualities of audio1", "tags": [ "voice-cloning", "vocal-consistency", "audio-reference", "snippet" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "references": "1 audio (~15s sample)" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-103", "model": "seedance-2.0", "title": "视频 reference 作为摄像机运动模板", "prompt": "use that video as a reference for camera movement", "tags": [ "video-reference", "camera-template", "motion-transfer", "snippet" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "references": "1 video (≤15s)" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-104", "model": "seedance-2.0", "title": "视频 reference 迁移角色动作(运镜解耦)", "prompt": "use this video as a character motion reference for image1", "tags": [ "video-reference", "motion-transfer", "character-motion", "decoupled", "snippet" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "references": "1 video + 1 image" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-105", "model": "seedance-2.0", "title": "替换视频中的角色(保留运镜+环境)", "prompt": "replace the character in the video with the dog", "tags": [ "video-reference", "character-replace", "scene-preserve", "snippet" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "references": "1 video + 1 image (new character)" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-106", "model": "seedance-2.0", "title": "风格+光影双迁移(视频参考)", "prompt": "using video1 as a style and lighting reference", "tags": [ "video-reference", "style-transfer", "lighting-transfer", "snippet" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "references": "1 video" }, "source": { "name": "YouTube TwcB2qX42iw · The Most Consistent AI Video Model Right Now", "url": "https://www.youtube.com/watch?v=TwcB2qX42iw", "tier": "reputable-3rd-party" } }, { "id": "sd-107", "model": "seedance-2.0", "title": "AI Anime 角色多角度参考图生成(送 Nano Banana)", "prompt": "Show this anime character in a unique setting, 90s anime style, no text, close-up, natural background", "tags": [ "anime", "character-sheet", "preprocessing", "nano-banana", "90s-style" ], "category": "creative", "params": { "aspect": "1:1", "duration": "n/a (image preprocessing)", "resolution": "n/a", "tool": "Nano Banana" }, "source": { "name": "YouTube MGfcx_NXRww · Seedance 2.0 Creates AI Anime With Perfect Character Consistency", "url": "https://www.youtube.com/watch?v=MGfcx_NXRww", "tier": "reputable-3rd-party" } }, { "id": "sd-108", "model": "seedance-2.0", "title": "Nano Banana Pro 3D 风格转换(绕过人脸过滤预处理)", "prompt": "turn this into a highly detailed 3D drawing art character design with white background concept art", "tags": [ "bypass-face", "nano-banana-pro", "3d-conversion", "preprocessing", "concept-art" ], "category": "creative", "params": { "aspect": "1:1", "duration": "n/a (image preprocessing)", "resolution": "n/a", "tool": "Nano Banana Pro" }, "source": { "name": "YouTube pbpzvmtpAZg · NEW Method for Consistent Characters Using Your OWN Face", "url": "https://www.youtube.com/watch?v=pbpzvmtpAZg", "tier": "reputable-3rd-party" } }, { "id": "sd-109", "model": "seedance-2.0", "title": "Ninja 热带海滩(3D→写实拉回照片质感)", "prompt": "Use character at image1. He's arriving to a tropical beach after his ship sank. 4K, photorealistic, live action, lifelike", "tags": [ "bypass-face", "3d-to-photoreal", "ninja", "beach", "live-action", "minimalist" ], "category": "creative", "params": { "aspect": "16:9", "duration": "5s", "resolution": "1080p", "references": "1 image (3D-converted character)" }, "source": { "name": "YouTube pbpzvmtpAZg · NEW Method for Consistent Characters Using Your OWN Face", "url": "https://www.youtube.com/watch?v=pbpzvmtpAZg", "tier": "reputable-3rd-party" } }, { "id": "sd-110", "model": "seedance-2.0", "title": "Transformers 巴黎变形 5 段多镜头(Image2+Image3 双参考)", "prompt": "[0-2 seconds] One-take, camera follows closely throughout. Scene begins on the banks of the Seine in Paris (Image 2), buildings collapsing, thick smoke, water currents roaring, Eiffel Tower faintly visible in smoke. [2-5 seconds] A male pilot in a white suit (Image 3) perilously straddles a giant mechanical structure. Camera zooms out, revealing he is riding a transforming fighter jet. Coat tails violently flapping in airflow as he tightly grips the mecha's components. [5-7 seconds] Fighter jet speeds close to Paris street level, body undergoing complex mechanical reconfiguration. Wings fold back, engines reorganize into humanoid mecha core, fuselage extending to form sturdy metal skeletons and glowing pipelines. [7-9 seconds] Transformation complete, giant mechanical behemoth stands in Paris streets. Machine raises head, back armor unfolds, revealing giant cannon barrel, energy accumulating. A devastating energy beam released with a roar, distant landmarks crumble into smoke. [9-10 seconds] Pilot climbs into cockpit. Mecha takes humanoid form, engines ignite at full power, twin exhaust flames intensely orange-red. Machine flies up into clouds, outline of Paris rapidly shrinking below.", "tags": [ "transformers", "mecha", "paris", "one-take", "multi-shot", "time-coded", "viral" ], "category": "creative", "params": { "aspect": "16:9", "duration": "10s", "resolution": "1080p", "references": "2 images (Image2 background + Image3 character)" }, "source": { "name": "YouTube pbpzvmtpAZg · NEW Method for Consistent Characters Using Your OWN Face", "url": "https://www.youtube.com/watch?v=pbpzvmtpAZg", "tier": "reputable-3rd-party" } }, { "id": "sd-111", "model": "seedance-2.0", "title": "POV 法师战场 — 一段描述自动多镜头(CapCut Video Studio)", "prompt": "A POV shot of a mage charging energy in a battlefield, forming a lightning sphere, then cutting to an opponent with a shield, and the impact creates a big explosion with particles", "tags": [ "pov", "mage", "battle", "auto-multi-shot", "lightning", "capcut-video-studio" ], "category": "creative", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p", "platform": "CapCut Video Studio" }, "source": { "name": "YouTube CfenG1mcb7c · I Just Tested Dreamina Seedance 2.0 in CapCut Video Studio", "url": "https://www.youtube.com/watch?v=CfenG1mcb7c", "tier": "reputable-3rd-party" } }, { "id": "sd-112", "model": "seedance-2.0", "title": "双猫功夫对决 — 港式复古武打风格(无时间戳多镜头)", "prompt": "Two cats kung fu fighting each other. They are wearing kung fu clothes. Hong Kong retro kung fu movie style. First cat says \"It's time to meet your end.\" Second cat replies \"Come at me, bro.\" They start fighting each other. Then the first cat hits the second cat with a hadouken. Second cat dodges and counters with a shoryuken.", "tags": [ "cats", "kung-fu", "hong-kong-retro", "hadouken", "dialogue", "viral", "auto-multi-shot" ], "category": "creative", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p", "platform": "CapCut Video Studio" }, "source": { "name": "YouTube 4zjPITR553w · How to Use Seedance 2.0 in CapCut - First Impressions", "url": "https://www.youtube.com/watch?v=4zjPITR553w", "tier": "reputable-3rd-party" } }, { "id": "sd-113", "model": "seedance-2.0", "title": "Fantasy 骑士 3 镜头时间轴(Advanced Guide 五段式示例)", "prompt": "[0-5s] Wide shot, majestic knight in silver armor enters a dark cave with a torch, ethereal god rays filtering from entrance. Hard cut. [5-10s] Close-up, knight's nervous eyes reflecting firelight. Subtle arc shot, shallow depth of field. Fade in. [10-15s] Medium tracking shot, low-angle. He draws a sword glowing blue. Sound of crackling fire and echoing footsteps.", "tags": [ "fantasy", "knight", "cave", "time-coded", "five-part-spine", "advanced-guide" ], "category": "creative", "params": { "aspect": "16:9", "duration": "15s", "resolution": "1080p" }, "source": { "name": "YouTube pbpzvmtpAZg · Advanced Prompting Guide (Google Doc)", "url": "https://www.youtube.com/watch?v=pbpzvmtpAZg", "tier": "reputable-3rd-party" } }, { "id": "sd-114", "model": "seedance-2.0", "title": "通用负向 Prompt 模板(Advanced Guide 4 类约束)", "prompt": "No jitter, no wobble, no warping, no temporal flicker. No added props, no color shift, no logo morphing. No text overlays, no garbled text. No face changes, no flicker, high consistency, keep the same character, same clothing, same hairstyle.", "tags": [ "negative-prompt", "constraints", "consistency", "template", "advanced-guide" ], "category": "technique-snippet", "params": { "aspect": "any", "duration": "any", "resolution": "any", "usage": "append to end of any Seedance prompt" }, "source": { "name": "YouTube pbpzvmtpAZg · Advanced Prompting Guide (Google Doc)", "url": "https://www.youtube.com/watch?v=pbpzvmtpAZg", "tier": "reputable-3rd-party" } }, { "id": "sd-115", "model": "seedance-2.0", "title": "⭐ 玩法 · 画线 FPV 航拍通用模板(MrLarus 原文)", "prompt": "请擦除红线、箭头和所有辅助标记。红线和箭头仅作为镜头运动路径参考,最终成片中不得出现。镜头以第一人称 FPV 视角呈现,超高速、电影级、一镜到底,严格沿着图片中的红色路径运动,不要偏离、不要跳步、不要简化路线。镜头从【起点】开始,经过【节点1】,随后【节点2】,再到【节点3】,最后到达【终点主体】并完成【收尾动作】。画面要求超写实、运动顺滑稳定、速度感强、空间连续清晰,不要重复建筑、不要变形、不要文字、不要水印。", "tags": [ "playbook", "fpv", "path-drawing", "drone-shot", "template", "image-to-video", "viral-technique" ], "category": "creative", "params": { "aspect": "16:9 或 9:16", "duration": "10-15s", "resolution": "1080p" }, "source": { "name": "@MrLarus(Larus Canus) · X tweet 评论区 — 画线控制 AI 航拍镜头玩法 (45K 浏览 · 275 赞)", "url": "https://x.com/MrLarus/status/2060588327409959082", "video_url": "https://video.twimg.com/amplify_video/2060586925543559171/vid/avc1/1080x1216/TO9DT93zSB4ChCJe.mp4?tag=27", "tier": "reputable-3rd-party", "note": "玩法集合 #1 — 在图片画红线作为飞行路径,Seedance 沿线生成 FPV 镜头。把【起点/节点/终点/收尾】换成自己场景即可。详见 methodology/21-fpv-航拍路径绘制玩法.md", "author": "@MrLarus (Larus Canus) — 工作流+prompt 模板由其在 X 公开分享" } }, { "id": "sd-116", "model": "seedance-2.0", "title": "玩法 · 世界地标大乱斗 FPV(富士山→埃菲尔→罗马→自由女神)", "prompt": "请擦除红线、箭头和所有辅助标记。红线和箭头仅作为镜头运动路径参考,最终成片中不得出现。镜头以第一人称 FPV 视角呈现,超高速、电影级、一镜到底,严格沿着图片中的红色路径运动,不要偏离、不要跳步、不要简化路线。镜头从富士山山顶开始,贴近富士山雪线掠过,经过埃菲尔铁塔尖顶,随后穿过罗马斗兽场圆形竞技场内部,再到纽约自由女神像火炬旁,最后绕火炬上升一周后定格在女神冠冕特写。画面要求超写实、运动顺滑稳定、速度感强、空间连续清晰,不要重复建筑、不要变形、不要文字、不要水印。", "tags": [ "playbook", "fpv", "landmarks-mashup", "world-tour", "image-to-video", "viral-technique" ], "category": "creative", "params": { "aspect": "9:16", "duration": "15s", "resolution": "1080p" }, "source": { "name": "@MrLarus · World Landmark Mashup FPV(主案例,45K 浏览)", "url": "https://x.com/MrLarus/status/2060588327409959082", "video_url": "https://video.twimg.com/amplify_video/2060586925543559171/vid/avc1/1080x1216/TO9DT93zSB4ChCJe.mp4?tag=27", "tier": "reputable-3rd-party", "note": "玩法集合 #1 实际填充案例。Workflow:画线图 + 此 prompt → Seedance", "author": "@MrLarus (Larus Canus) — 工作流+prompt 模板由其在 X 公开分享" } }, { "id": "sd-117", "model": "seedance-2.0", "title": "玩法 · 广州中轴线 FPV(花城广场→楼顶→珠江→小蛮腰)", "prompt": "请擦除红线、箭头和所有辅助标记。红线和箭头仅作为镜头运动路径参考,最终成片中不得出现。镜头以第一人称 FPV 视角呈现,超高速、电影级、一镜到底,严格沿着图片中的红色路径运动,不要偏离、不要跳步、不要简化路线。镜头从花城广场低空开始,贴楼拔高沿着 CBD 玻璃幕墙上升,随后穿过楼顶结构与天台,接着绕楼俯冲掠过珠江水面,再冲向广州塔小蛮腰,最后绕塔螺旋上升一周后定格在塔顶。画面要求超写实、运动顺滑稳定、速度感强、空间连续清晰,城市建筑细节真实,不要重复建筑、不要变形、不要文字、不要水印。", "tags": [ "playbook", "fpv", "city-aerial", "guangzhou", "image-to-video", "real-estate", "viral-technique" ], "category": "creative", "params": { "aspect": "9:16", "duration": "15s", "resolution": "1080p" }, "source": { "name": "@MrLarus · 广州中轴线 FPV(玩法起源,158K 浏览 · 1389 赞 · 210 RT)", "url": "https://x.com/MrLarus/status/2060267899382542341", "video_url": "https://video.twimg.com/amplify_video/2060267899382542341/vid/avc1/1080x1216/I__iW70_wy-nHwIg.mp4?tag=27", "tier": "reputable-3rd-party", "note": "玩法集合 #1 城市航拍案例。可换任何城市轴线:北京中轴线 / 上海外滩 / 香港维港 / 巴黎香榭大道 等", "author": "@MrLarus (Larus Canus) — 工作流+prompt 模板由其在 X 公开分享" } }, { "id": "so-001", "model": "sora-2", "category": "cinematic", "title": "机械工坊小机器人(OpenAI 官方 Shot List 样板)", "prompt": "Style: Hand-painted 2D/3D hybrid animation with soft brush textures, warm tungsten lighting, and tactile stop-motion feel. Mid-2000s storybook animation aesthetic—cozy, imperfect, mechanical charm. Watercolor wash, painterly textures, warm–cool balance, filmic motion blur.\n\nInside cluttered workshop with gear-laden shelves and blueprints. Small round robot sits on wooden bench, dented body with mismatched patches. Large glowing blue eyes flicker as it fiddles with humming light bulb.\n\nCinematography: Medium close-up, slow push-in with parallax from hanging tools. 35mm virtual lens, shallow DOF softening background. Warm overhead practical key; cool window spill. Gentle, whimsical, slightly suspenseful mood.\n\nActions:\n- Robot taps bulb; sparks crackle\n- Flinches, dropping bulb, eyes widen\n- Bulb tumbles in slow motion; catches it\n- Steam escapes chest—relief and pride\n- Robot says: \"Almost lost it… but I got it!\"\n\nBackground Sound: Rain, ticking clock, soft mechanical hum, faint bulb sizzle.", "tags": [ "sora-cookbook", "animation", "shot-list", "dialogue", "audio" ], "params": { "aspect": "16:9", "duration": "4s" }, "source": { "name": "OpenAI Cookbook", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Sora 2 官方推荐分层结构:Style → Scene → Cinematography → Actions(按 beats)→ Sound" }, { "id": "so-002", "model": "sora-2", "category": "cinematic", "title": "70 年代屋顶浪漫(OpenAI 官方)", "prompt": "Style: 1970s romantic drama, shot on 35mm film with natural flares, soft focus, warm halation. Slight gate weave and handheld micro-shake evoke vintage intimacy. Warm Kodak-inspired grade; light halation on bulbs; film grain and soft vignette for period authenticity.\n\nAt golden hour, brick tenement rooftop transforms into small stage. Laundry lines strung with white sheets sway in wind. Mismatched fairy bulbs hum overhead. Young woman in flowing red silk dress dances barefoot, curls glowing. Partner—sleeves rolled, suspenders loose—claps along.\n\nCinematography: Medium-wide shot, slow dolly-in from eye level. 40mm spherical lens, shallow focus isolating couple from skyline. Golden natural key with tungsten bounce; edge from fairy bulbs. Nostalgic, tender, cinematic mood.\n\nActions:\n- She spins; dress flares catching sunlight\n- Woman (laughing): \"See? Even the city dances with us tonight.\"\n- He steps in, catches her hand, dips her into shadow\n- Man (smiling): \"Only because you lead.\"\n- Sheets drift across frame, briefly veil skyline before parting\n\nBackground Sound: Natural ambience only—wind, fabric flutter, street noise, muffled music. No added score.", "tags": [ "sora-cookbook", "period", "romance", "35mm", "dialogue" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "OpenAI Cookbook", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "黄金时段+对白+实景拍片质感的官方样板,Style 段锁定胶片美学" }, { "id": "so-003", "model": "sora-2", "category": "cinematic", "title": "通勤站台超精细参数化(OpenAI 官方)", "prompt": "Format & Look: Duration 4s; 180° shutter; digital capture emulating 65mm photochemical contrast; fine grain; subtle halation on speculars; no gate weave.\n\nLenses & Filtration: 32mm / 50mm spherical primes; Black Pro-Mist 1/4; slight CPL rotation to manage glass reflections on train windows.\n\nGrade/Palette: Highlights—clean morning sunlight with amber lift. Mids—balanced neutrals with slight teal cast in shadows. Blacks—soft, neutral with mild lift.\n\nLighting & Atmosphere: Natural sunlight from camera left (07:30 AM). 4×4 ultrabounce silver; negative fill from opposite wall. Gentle mist; train exhaust drift.\n\nLocation & Framing: Urban commuter platform at dawn. Foreground—yellow safety line, coffee cup. Midground—silhouetted passengers in haze. Background—arriving train braking.\n\nWardrobe: Mid-30s traveler, navy coat, backpack, holding phone loosely.\n\nSound: Diegetic only—rail screech, train brakes, muffled announcement, ambient hum, footsteps, paper rustle.", "tags": [ "sora-cookbook", "ultra-detailed", "cinematography", "technical" ], "params": { "aspect": "21:9", "duration": "4s" }, "source": { "name": "OpenAI Cookbook", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Sora 2 的「制作蓝皮书」级提示词模板:Format → Lenses → Grade → Lighting → Location → Wardrobe → Sound" }, { "id": "so-004", "model": "sora-2", "category": "dialogue-driven", "title": "审讯室对白(OpenAI 官方 Dialogue 块)", "prompt": "Style: Neo-noir crime drama, 35mm anamorphic, deep shadows, single-source key light, smoke-thick atmosphere, desaturated palette with bruised teal blacks.\n\nA cramped, windowless room with walls the color of old ash. A single bare bulb dangles from the ceiling, its light pooling onto the scarred metal table at the center. Two chairs face each other across it. On one side sits the Detective, trench coat draped across the back of his chair, eyes sharp and unblinking. Across from him, the Suspect slouches, cigarette smoke curling lazily toward the ceiling. The silence presses in.\n\nCinematography:\n- Camera: medium two-shot, locked off; subtle push-in on the Detective for the final line\n- Lens: 40mm anamorphic; shallow focus drifting between faces\n- Lighting: hard top-down practical, no fill; smoke catches the beam\n- Mood: claustrophobic, accusatory\n\nDialogue:\n- Detective: \"You're lying. I can hear it in your silence.\"\n- Suspect: \"Or maybe I'm just tired of talking.\"\n- Detective: \"Either way, you'll talk before the night's over.\"\n\nBackground Sound: Faint hum of the overhead light; chair creak; one slow exhale of smoke.", "tags": [ "sora-cookbook", "dialogue", "noir", "interrogation", "audio" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "OpenAI Cookbook — Dialogue and Audio", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 示范如何把 Dialogue 单独成块、用 alternating turns 让模型把每行对应到正确角色" }, { "id": "so-005", "model": "sora-2", "category": "documentary", "title": "90 年代纪录片访谈(短提示词版)", "prompt": "Style: 90s documentary interview, 16mm with soft halation, slight gate weave, warm Kodachrome grade, fine grain.\n\nIn a wood-paneled study lined with old books and a single lamp, an old Swedish man—weathered hands, wool cardigan, kind eyes—sits in a worn leather armchair. Afternoon light slants through linen curtains. He looks slightly off-camera as if speaking to an unseen interviewer.\n\nCinematography:\n- Camera: medium close-up, eye level, locked off on a small tripod\n- Lens: 50mm vintage prime; gentle shallow focus on his face\n- Lighting: single window key from camera left; warm practical fill from desk lamp\n- Mood: reflective, intimate, unhurried\n\nActions:\n- He folds his hands slowly in his lap\n- Old Swedish man (softly): \"I still remember when I was young.\"\n- A small smile, eyes drifting toward the window\n\nBackground Sound: Faint clock ticking; distant traffic hum; one page settling on the desk. No music.", "tags": [ "sora-cookbook", "short-prompt", "documentary", "interview", "16mm" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "OpenAI Cookbook — Short prompt example", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 原版只有一句话;这里按官方说法把它扩成 Shot-List 仍保留留白让模型发挥" }, { "id": "so-006", "model": "sora-2", "category": "cinematic", "title": "霓虹湿街强写法对比(Cookbook Strong Prompt)", "prompt": "Style: Contemporary urban cinematic, anamorphic 2.0x, shallow DOF, volumetric light through humid air, neon-saturated palette of magenta, cyan, and amber.\n\nA narrow city street after rain. Wet asphalt with painted zebra crosswalk reflecting overhead signs. Neon storefront signs—red, pink, cyan—shimmer in shallow puddles. Steam drifts from a manhole. The sidewalk is empty except for a lone figure waiting at the curb.\n\nCinematography:\n- Camera: low-angle wide, eye level with the pavement, slow push-in along the crosswalk lines\n- Lens: 35mm anamorphic; horizontal lens flares from neon\n- Lighting: neon signs as colored key; underlit by reflections in wet asphalt\n- Mood: cinematic, expectant, lonely\n\nActions:\n- A taxi rolls past slowly; reflections fracture in the puddle\n- The figure steps once toward the curb and stops\n- A traffic light cycles from red to green; the puddle ripples\n\nBackground Sound: Distant traffic hiss; tire spray on wet road; faint hum from a flickering sign. No score.", "tags": [ "sora-cookbook", "strong-prompt", "anamorphic", "neon", "rain" ], "params": { "aspect": "21:9", "duration": "8s" }, "source": { "name": "OpenAI Cookbook — Weak vs Strong prompts", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 用「wet asphalt, zebra crosswalk, neon sign reflection」+「anamorphic 2.0x, shallow DOF, volumetric light」作为 strong prompt 样板;本条把它扩成完整 Shot-List" }, { "id": "so-007", "model": "sora-2", "category": "cinematic", "title": "骑行者刹停十字路口(Beats > Seconds)", "prompt": "Style: Naturalistic documentary, handheld ENG camera, slight rolling shutter wobble, warm afternoon grade with cool shadows.\n\nA quiet residential intersection. Late-afternoon sun rakes across the asphalt. Trees throw long, dappled shadows on the crosswalk. A young cyclist in a faded windbreaker approaches from frame left, breath visible in the cool air.\n\nCinematography:\n- Camera: handheld eye-level, tracks the cyclist briefly from the side, then settles\n- Lens: 28mm; deep focus keeping street and rider sharp\n- Lighting: golden hour sun from camera right; cool blue fill from sky\n- Mood: grounded, observational, gently kinetic\n\nActions:\n- Cyclist pedals three times, coasting toward the crosswalk\n- Brakes squeeze; the bike slows, rear wheel barely chirps on asphalt\n- Foot drops; cyclist stops fully at the crosswalk, glances right\n- A small exhale, then steady stillness for the final beat\n\nBackground Sound: Bike-chain tick; brake-pad squeak; distant lawnmower; one bird. No music.", "tags": [ "sora-cookbook", "beats", "timing", "handheld", "naturalistic" ], "params": { "aspect": "16:9", "duration": "4s" }, "source": { "name": "OpenAI Cookbook — Control motion and timing", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 原句:「Cyclist pedals three times, brakes, and stops at crosswalk」—— 用可数 beats 替代「moves quickly」" }, { "id": "so-008", "model": "sora-2", "category": "quiet-moments", "title": "四步到窗前拉开窗帘(精确动作分解)", "prompt": "Style: Quiet drama, naturalistic, soft window-light cinematography, muted earth-tone palette of amber, cream, walnut brown.\n\nA narrow apartment bedroom at late morning. A wooden floor, an unmade bed in soft shadow, sheer curtains drawn shut over a tall window. Pale light bleeds through the fabric. An actor in a simple linen shirt stands a few paces from the window, hesitating.\n\nCinematography:\n- Camera: medium-wide, locked off at eye level; one slow micro-push toward the window\n- Lens: 50mm; shallow DOF, sharp on the actor\n- Lighting: soft window light as key; warm lamp fill on the bed; cool rim from hallway\n- Mood: contemplative, tender, quietly hopeful\n\nActions:\n- Actor takes four steps to the window\n- Pauses, hand hovering near the curtain\n- Inhales once\n- Pulls the curtain open in the final second; daylight floods the room\n\nBackground Sound: Floorboard creak under each step; fabric slide of the curtain; distant city hum beyond the glass. No music.", "tags": [ "sora-cookbook", "beats", "precise-timing", "lighting", "intimate" ], "params": { "aspect": "16:9", "duration": "4s" }, "source": { "name": "OpenAI Cookbook — Beats and lighting examples", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "组合 Cookbook 两条原文:1) 「Actor takes four steps to the window, pauses, and pulls the curtain in the final second」;2) 「soft window light with warm lamp fill, cool rim from hallway / palette anchors: amber, cream, walnut brown」" }, { "id": "so-009", "model": "sora-2", "category": "quiet-moments", "title": "她转身微笑然后离开画面(图像输入参考)", "prompt": "Style: Soft contemporary portrait cinema, naturalistic skin tones, gentle film grain, painterly background blur.\n\n[Use image input as first-frame reference: a young woman standing in soft afternoon light, three-quarter angle, environment locked in by the image.]\n\nCinematography:\n- Camera: locked off medium close-up matching the reference frame; very subtle handheld breathing\n- Lens: 50mm; shallow DOF preserving the reference's separation between subject and background\n- Lighting: maintain the reference's soft directional key and ambient fill\n- Mood: warm, unguarded, momentary\n\nActions:\n- She turns around\n- A small, genuine smile reaches her eyes\n- She holds the look for one beat\n- She slowly walks out of the frame to camera right; light continues to play on the empty background\n\nBackground Sound: Quiet room tone; faint cloth rustle as she turns; a single distant door. No music.\n\nAPI note: pass the reference image via input_reference; size must match the target video resolution (jpeg/png/webp).", "tags": [ "sora-cookbook", "image-input", "reference", "portrait", "api" ], "params": { "aspect": "9:16", "duration": "4s" }, "source": { "name": "OpenAI Cookbook — Use image input for more control", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 原始 image-input prompt:「She turns around and smiles, then slowly walks out of the frame.」—— 这里加上官方推荐的 Shot-List 包装" }, { "id": "so-010", "model": "sora-2", "category": "comedy", "title": "冰箱门里钻出毛绒紫色怪物(图像输入参考)", "prompt": "Style: Pixar-adjacent stylized 3D animation, soft global illumination, plush fabric and squash-and-stretch motion, mid-bright kitchen palette.\n\n[Use image input as first-frame reference: a tidy kitchen with a closed refrigerator front and center; lighting and set dressing locked by the image.]\n\nCinematography:\n- Camera: locked off wide-medium, eye level with the fridge handle; one tiny rack focus when the door opens\n- Lens: 35mm; deep enough focus to keep both fridge and floor crisp\n- Lighting: warm overhead kitchen practical; cool blue spill from inside the fridge once it opens\n- Mood: playful, slightly mischievous, family-friendly\n\nActions:\n- The fridge door opens on its own with a soft pneumatic sigh\n- Cold blue light pours out across the floor\n- A cute, chubby purple monster waddles out, blinks twice, and tilts its head at the camera\n- It hugs a cartoon carton of milk to its chest and grins\n\nBackground Sound: Fridge seal pop; soft hum of compressor; tiny waddle squeaks. No score.", "tags": [ "sora-cookbook", "image-input", "animation", "character", "comedy" ], "params": { "aspect": "9:16", "duration": "4s" }, "source": { "name": "OpenAI Cookbook — Use image input for more control", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 原始 image-input prompt:「The fridge door opens. A cute, chubby purple monster comes out of it.」 + 官方建议的 Shot-List 扩写" }, { "id": "so-011", "model": "sora-2", "category": "action", "title": "金毛 Alfie 日出狂奔(Characters API 一致性)", "prompt": "Style: Cinematic outdoor naturalism, slow-motion-ready, golden-hour grade with cool morning shadows, fine grain.\n\nA wide meadow at sunrise, wet grass beaded with dew, low fog drifting between the blades. The horizon line catches first light, painting the sky pale gold and salmon. Alfie, a buoyant golden retriever, charges across frame from camera right toward camera left.\n\nCinematography:\n- Camera: low-angle tracking shot at knee height, dollying with Alfie\n- Lens: 35mm; deep focus to keep meadow and dog sharp\n- Lighting: warm rim from rising sun camera left; cool fill from open sky\n- Mood: joyful, free, kinetic\n\nActions:\n- Alfie bounds three strides, ears flapping, tongue out\n- Splashes through a shallow puddle; droplets arc in slow motion\n- A burst of fog parts around his shoulders\n- He veers slightly toward camera, eyes bright, then continues past\n\nBackground Sound: Paws on wet grass; water spray; soft dawn birdsong. No score.\n\nAPI: model: sora-2, size 720x1280, seconds 4, characters [{ id: char_alfie }] — name Alfie in the prompt so the model binds the reference identity.", "tags": [ "sora-cookbook", "characters-api", "consistency", "animal", "outdoor" ], "params": { "aspect": "9:16", "duration": "4s" }, "source": { "name": "OpenAI Cookbook — Characters API example", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 原 JSON 样例 prompt:「A cinematic shot of Alfie running through wet grass at sunrise」—— 这里展开成完整 Shot-List,保留 Characters API 用法" }, { "id": "so-012", "model": "sora-2", "category": "cinematic", "title": "IMAX 航拍冰川史诗(Style 锚定法)", "prompt": "Style: Epic, IMAX-scale aerial cinematography, 70mm large-format texture, ultra-clear deep blacks, restrained color grade favoring glacial blues and pale gold sunlight.\n\nA vast tidewater glacier meets a still arctic fjord at dawn. Cracked ice fields stretch toward distant snow-capped peaks. A faint mist rises off open water; a lone iceberg drifts mid-frame. No people, no boats.\n\nCinematography:\n- Camera: aerial wide shot, slight downward angle, slow forward push toward the glacier face\n- Lens: 24mm spherical; deep focus from foreground floes to distant peaks\n- Lighting: low-angle sunrise key from camera right; cool ambient fill from open sky\n- Mood: monumental, hushed, geological\n\nActions:\n- Camera glides forward at steady altitude; an ice shelf calves silently in the distance\n- The chunk slips into the sea, sending a slow, wide ripple toward camera\n- The iceberg rolls a quarter turn, revealing translucent blue underbelly\n\nBackground Sound: Distant low ice groan; wind across open water; one sharp crack at calving. No music.", "tags": [ "sora-cookbook", "style-anchor", "imax", "aerial", "landscape" ], "params": { "aspect": "21:9", "duration": "8s" }, "source": { "name": "OpenAI Cookbook — Visual cues that steer the look", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 把「epic, IMAX-scale scene」「aerial wide shot, slight downward angle」列为 style anchor 与 framing 示例 —— 此条把两者合并成完整提示词" }, { "id": "so-013", "model": "sora-2", "category": "documentary", "title": "16mm 黑白手持街头纪录片", "prompt": "Style: Vintage 16mm black-and-white documentary, high-contrast silver-gelatin grade, visible grain, slight gate weave, mild lens vignette.\n\nA narrow market street in a working-class neighborhood, mid-morning, late autumn. Vendors set up wooden stalls; steam rises from a kettle. Pedestrians cross frame in soft motion blur. A young vendor in a wool cap arranges apples into a pyramid.\n\nCinematography:\n- Camera: handheld ENG camera at chest height; gentle drift between subjects, no zooms\n- Lens: 25mm prime; deep focus from foreground crates to background storefront\n- Lighting: overcast natural sky as soft key; no fill; deep blacks in stall shadows\n- Mood: observational, warm despite monochrome, lived-in\n\nActions:\n- Vendor lifts an apple, polishes it on his apron, places it atop the pyramid\n- A child runs past chasing a paper bag\n- Vendor glances up at the camera, half-smiles, returns to stacking\n\nBackground Sound: Crate scrape; coins clinking; nearby radio playing a faint folk tune; footsteps on cobble. No score.", "tags": [ "sora-cookbook", "style-anchor", "16mm", "black-and-white", "handheld" ], "params": { "aspect": "4:3", "duration": "8s" }, "source": { "name": "OpenAI Cookbook — Visual cues that steer the look", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 提到「vintage 16mm documentary」与「handheld eng camera」作为风格锚点;本条按官方 anatomy(style → scene → cinematography → actions → sound)完整组装" }, { "id": "so-014", "model": "sora-2", "category": "quiet-moments", "title": "暖窗光与冷走廊光的厨房早晨(Lighting Recipe)", "prompt": "Style: Quiet domestic realism, naturalistic color science, gentle film emulation, palette anchors: amber, cream, walnut brown.\n\nA small kitchen at 8:00 AM. Tall window over the sink, sheer curtain half-open. A walnut breakfast table holds a single ceramic mug, a folded newspaper, and a bowl of cream-colored linen napkins. Hallway behind the table opens to a cooler, bluer adjoining room.\n\nCinematography:\n- Camera: medium-wide, locked off; one almost-imperceptible micro-push toward the mug\n- Lens: 40mm; shallow DOF favoring the mug and newspaper\n- Lighting: soft window light as key from camera left; warm lamp fill on the table; cool rim from the hallway\n- Mood: calm, lived-in, hopeful\n\nActions:\n- Steam rises in a slow ribbon from the mug\n- A draft lifts the curtain edge a few inches and lets it drop\n- A hand enters frame, picks up the mug, lifts it out of frame; the saucer remains\n\nBackground Sound: Refrigerator hum; clock ticking; faint kettle settling in the next room. No music.", "tags": [ "sora-cookbook", "lighting-recipe", "palette-anchors", "domestic", "still-life" ], "params": { "aspect": "16:9", "duration": "4s" }, "source": { "name": "OpenAI Cookbook — Lighting and color consistency", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 强提示样板:「Lighting + palette: soft window light with warm lamp fill, cool rim from hallway / Palette anchors: amber, cream, walnut brown」—— 完整 Shot-List 化" }, { "id": "so-015", "model": "sora-2", "category": "dialogue-driven", "title": "咖啡馆早午餐对话(Audio 与 Dialogue 节奏)", "prompt": "Style: Contemporary indie drama, naturalistic, 35mm Super 35 emulation, warm afternoon light with subtle haze, gently desaturated palette.\n\nA neighborhood café mid-afternoon. A two-top by the window; light through bistro curtains scatters across a marble tabletop. Two friends in their late 20s sit across from one another over half-finished lattes. One leans on her elbow; the other turns a spoon idly.\n\nCinematography:\n- Camera: shoulder-level shot/reverse-shot is implied but stay on a single over-the-shoulder framing on Speaker A; very small handheld breathing\n- Lens: 50mm; shallow DOF, sharp on faces\n- Lighting: warm window backlight as key; gentle bounce on the inside cheek; cool ambient fill from the café interior\n- Mood: candid, intimate, slightly nervous\n\nActions:\n- Speaker A (quietly): \"I almost didn't come today.\"\n- Speaker B (gentle): \"I know.\"\n- Speaker A glances down at her cup, traces the rim with one finger\n- Speaker B (softly): \"I'm glad you did.\"\n\nBackground Sound: The hum of espresso machines and the murmur of voices form the background; cup tap on saucer; one chair scrape across the room. No score.", "tags": [ "sora-cookbook", "dialogue", "naturalistic", "two-shot", "audio" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "OpenAI Cookbook — Dialogue and Audio", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 示范的环境声原句:「The hum of espresso machines and the murmur of voices form the background.」+ 官方建议 4s 容 1-2 句对白、8s 容更多" }, { "id": "so-016", "model": "sora-2", "category": "cinematic", "title": "迭代式镜头切换 85mm(Edit Prompt 工作流)", "prompt": "Style: Continuation of a previous Sora 2 generation. Maintain the locked aesthetic — same wardrobe, same set dressing, same color grade. Only change what is specified below.\n\nReference: same subject, same location, same lighting setup as the original clip.\n\nCinematography (changes only):\n- Same shot, switch to 85mm\n- Shallower depth of field; tighter framing on the subject's eyes and hands\n- Keep camera height and angle identical to the original\n\nActions (unchanged):\n- Subject performs the original action exactly as before; do not re-block, do not re-light\n\nBackground Sound: Identical to the original — diegetic only, no added score.\n\nProduction note: this is an EDIT prompt, not a fresh generation. Pin the original as reference and describe only the change (\"same shot, switch to 85mm\"). Sora 2 Edit endpoint: POST /v1/videos/{video_id}/edits.", "tags": [ "sora-cookbook", "edit-endpoint", "iteration", "workflow", "api" ], "params": { "aspect": "16:9", "duration": "4s" }, "source": { "name": "OpenAI Cookbook — Iterate with video edits", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 原文:「same shot, switch to 85 mm」「Editing is for nudging, not gambling」—— 这条把 Edit Workflow 直接转成可复用模板" }, { "id": "so-017", "model": "sora-2", "category": "horror", "title": "无人长走廊缓推(Style + 最小动作)", "prompt": "Style: Slow-burn psychological horror, 1.85:1 framing, 35mm anamorphic, desaturated palette of bone, pale green, and shadow black, fine grain, restrained halation.\n\nA long, empty hospital corridor at night. Linoleum floor reflects a single flickering fluorescent tube halfway down. Closed doors line both walls; numbers stenciled in faded paint. The far end of the corridor disappears into darkness.\n\nCinematography:\n- Camera: locked off centered on the vanishing point, very slow dolly-in along the corridor axis\n- Lens: 32mm anamorphic; deep focus the whole length\n- Lighting: single overhead fluorescent as key; cool ceiling spill; deep blacks at the far end\n- Mood: dread, restraint, anticipation\n\nActions:\n- The fluorescent tube buzzes once and dims for half a beat\n- A door at the far end clicks softly but does not move\n- The camera holds for the final beat; nothing more happens\n\nBackground Sound: Fluorescent ballast hum; distant ventilation whine; one quiet click from the door. No score, no jump scare.", "tags": [ "sora-cookbook", "horror", "restraint", "atmospheric", "anamorphic" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "OpenAI Cookbook — Style is the strongest lever", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "对照 Cookbook 反复强调的「Establish style early」+「Each shot should have one clear camera move and one clear subject action」—— 用克制的两个 beat 替代「scary jump scare」式描述" }, { "id": "so-018", "model": "sora-2", "category": "cinematic", "title": "多镜头连续性两段(Multi-shot 拼接)", "prompt": "Style: Naturalistic indie drama, 35mm spherical, soft afternoon grade, palette anchors: oat, slate blue, soft rust. Keep style identical across both shots so they cut cleanly.\n\nShot 1 (0.0–4.0s) — Apartment doorway:\n- Camera: medium-wide, locked off at eye level\n- Lens: 35mm; shallow DOF on the subject\n- Lighting: warm hallway key from camera right, cool spill from open apartment\n- Action: A woman in a soft rust sweater pauses in the doorway, sets a paper grocery bag on the floor, looks back over her shoulder once\n\nShot 2 (4.0–8.0s) — Kitchen counter:\n- Camera: medium close-up, locked off\n- Lens: 50mm; same shallow DOF; same wardrobe and lighting logic\n- Lighting: window light from camera left, warm pendant fill above the counter\n- Action: Same woman unpacks one apple from the bag, places it on a wooden board, exhales once\n\nBackground Sound: Door click between shots; bag rustle; apple set down softly. No score, no transition sweetener.\n\nNote: each shot has one camera setup, one action, one lighting recipe — per Cookbook's multi-shot guidance. Stitch the two clips in edit or let them play as a sequence.", "tags": [ "sora-cookbook", "multi-shot", "continuity", "two-shot-sequence", "edit-ready" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "OpenAI Cookbook — Multi-shot prompts", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 原句:「Describing multiple shots in a single prompt is also valid... keep each shot block distinct: one camera setup, one subject action, and one lighting recipe at a time.」" }, { "id": "so-019", "model": "sora-2", "category": "action", "title": "东京霓虹巷雨夜步行(Atlabs 7 层结构)", "prompt": "Style: Contemporary cinematic, anamorphic 2.0x, shallow DOF, volumetric light through wet air, neon-saturated palette of magenta, cyan, amber.\n\nScene Setup: A bustling neon-lit Tokyo alley at midnight during a light rain, with reflective puddles on the pavement.\n\nSubject and Action: A young woman in a leather jacket walks confidently, glancing over her shoulder with determination.\n\nCinematography:\n- Camera: wide-angle, tracking shot from behind, low to the ground\n- Lens: 24mm; shallow DOF blurring the background crowds\n- Lighting: neon signs provide vibrant key with cool blue and pink hues, casting glowing reflections on wet surfaces\n- Mood: confident, kinetic, slightly mysterious\n\nActions:\n- She walks three steady steps; the puddle ripples\n- Glances over her shoulder once toward camera\n- Continues forward; a passing umbrella crosses the foreground\n\nPhysics & Materials: Rain droplets ripple in puddles; wind gently rustles the jacket's fabric; realistic shadows from overhead lights.\n\nBackground Sound: Soft rain patter and distant city hum; footsteps splashing; dialogue: \"Time to move.\"\n\nExclusions: No legible text on signs; avoid lens flares or unnatural colors.", "tags": [ "atlabs", "shot-list", "neon", "rain", "tokyo" ], "params": { "aspect": "21:9", "duration": "8s" }, "source": { "name": "Atlabs AI — Sora 2 Prompt Guide (Best Practices 2026)", "url": "https://www.atlabs.ai/blog/sora-2-prompt-guide", "tier": "reputable-3rd-party" }, "notes": "Atlabs 7 层结构(Scene / Subject / Camera / Lighting / Physics / Audio / Exclusions)的官方完整样例,原文给出的每一层都直接引用" }, { "id": "so-020", "model": "sora-2", "category": "cinematic", "title": "怪物变橙色:编辑提示词(Edit Prompt 微调)", "prompt": "Style: Continuation of a previous Sora 2 generation. Keep every other variable identical to the source clip — composition, framing, motion, lighting, audio. Only the change below.\n\nReference: previous clip featuring a chubby cartoon monster.\n\nChange (single edit only):\n- Change the color of the monster to orange\n\nKeep unchanged:\n- Same shot, same framing, same camera move\n- Same wardrobe/props, same set, same lighting recipe\n- Same action beats, same dialogue (if any)\n- Same diegetic sound layers\n\nProduction note: this is a SURGICAL edit prompt, not a fresh generation. Cookbook guidance: \"Editing is for nudging, not gambling. Use it to make controlled changes — one at a time.\" Sora 2 Edit endpoint: POST /v1/videos/{video_id}/edits.\n\nBackground Sound: identical to the source clip.", "tags": [ "sora-cookbook", "edit-endpoint", "single-change", "workflow", "iteration" ], "params": { "aspect": "16:9", "duration": "4s" }, "source": { "name": "OpenAI Cookbook — Iterate with video edits", "url": "https://developers.openai.com/cookbook/examples/sora/sora2_prompting_guide", "tier": "official" }, "notes": "Cookbook 原 edit prompt:「Change the color of the monster to orange」+「A second monster comes out right after」—— 教用户怎么做最小变更迭代" }, { "id": "ve-001", "model": "veo-3", "category": "dialogue-driven", "title": "老水手的烟斗箴言(Veo 3 原生对白样板)", "prompt": "A medium shot frames an old sailor with a weathered face and a thick white beard. He holds his pipe in one hand and gestures towards the churning sea with the other. Dialogue: \"This ocean, it's a force, a wild, untamed might. And she commands respect at all times.\" Audio: Crash of waves, distant gull cries, the creaking of an old wooden boat under his feet, a low gust of wind.", "tags": [ "character", "dialogue", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Google DeepMind", "url": "https://deepmind.google/models/veo/prompt-guide/", "tier": "official" }, "notes": "Veo 3 的「角色+对白+多层环境音」官方公式:人物外观→动作→Dialogue:→Audio:" }, { "id": "ve-002", "model": "veo-3", "category": "dialogue-driven", "title": "糖果键盘(感官音效)", "prompt": "A keyboard whose keys are made of brightly colored candy, glistening under soft studio light. Hands type rapidly across the surface. Each keystroke produces sweet, crunchy candy-breaking sounds. Audio: Crunchy, sugary typing sounds, delighted giggles in the background, a faint pop song humming.", "tags": [ "whimsical", "audio", "sensory", "veo-official" ], "params": { "aspect": "1:1", "duration": "6s" }, "source": { "name": "Google DeepMind", "url": "https://deepmind.google/models/veo/prompt-guide/", "tier": "official" } }, { "id": "ve-003", "model": "veo-3", "category": "dialogue-driven", "title": "露营定格动画对白", "prompt": "Stop motion claymation style. Inside a small tent at twilight, a wide-eyed clay camper sits opposite a grumpy clay bear. Camper says cheerfully: \"I'm one with nature now!\" Bear deadpans: \"Nature would prefer personal space.\" Audio: Soft crackling campfire outside, distant owl, exaggerated claymation breath sounds.", "tags": [ "claymation", "comedy", "dialogue", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "Google DeepMind", "url": "https://deepmind.google/models/veo/prompt-guide/", "tier": "official" } }, { "id": "ve-004", "model": "veo-3", "category": "action", "title": "硬核越野拉力(多层音效)", "prompt": "An 8-second hardcore off-road rally sequence. Three mud-caked 4x4 vehicles power across a wide river crossing, water spraying violently from oversized tires. Wide tracking shot from the riverbank, then a low-angle hero shot following the lead vehicle. Audio: Roaring V8 engines, sloshing water, gravel impacts, tire chain rattles, occasional driver shouts from inside the cabin.", "tags": [ "action", "rally", "audio", "veo-official" ], "params": { "aspect": "21:9", "duration": "8s" }, "source": { "name": "Google DeepMind", "url": "https://deepmind.google/models/veo/prompt-guide/", "tier": "official" } }, { "id": "ve-005", "model": "veo-3", "category": "cinematic", "title": "1980s 加班族(怀旧胶片质感)", "prompt": "Retro cinematic aesthetic, shot as if on 1980s color film with subtle grain. Medium shot of a tired corporate worker in his late thirties, loosened tie and rolled-up shirt sleeves, rubbing his temples in exhaustion in front of a bulky beige 1980s CRT computer. The cluttered office is late at night, papers stacked beside a half-empty coffee mug. Static camera, no movement. Lighting: harsh fluorescent overhead tubes mixed with the green glow of the monochrome monitor washing over his face. Environment: dim cubicle, blinds half drawn, distant rain against the window. Audio: low hum of CRT and fluorescent buzz, soft mechanical keyboard taps in another room, faint AM radio playing a melancholy jazz tune. Dialogue: (under his breath) \"Just one more report... one more.\"", "tags": [ "cinematic", "retro", "1980s", "office", "audio", "dialogue", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud Blog — Ultimate Prompting Guide for Veo 3.1", "url": "https://cloud.google.com/blog/products/ai-machine-learning/ultimate-prompting-guide-for-veo-3-1", "tier": "official" }, "notes": "Google 官方「Retro Office Worker」示例改写为完整 8 元素 + 双标记。原型展示怀旧胶片质感 + 单一光源(显示器+荧光灯)。" }, { "id": "ve-006", "model": "veo-3", "category": "cinematic", "title": "雾谷孤行者(起重机镜头展示)", "prompt": "Epic fantasy cinematic style, awe-inspiring atmosphere. Subject: a lone hiker in a weathered green parka and heavy boots, backpack silhouetted against the sky, standing on a narrow rocky ledge. Action: hiker pauses and lifts gaze across the abyss. Camera: crane shot starting low at the hiker's feet and ascending high above to reveal a colossal mist-filled canyon stretching to the horizon at sunrise. Lighting: soft golden morning light skimming the canyon rim, deep blue shadows in the valley below, volumetric god-rays slicing through the mist. Environment: vast rocky canyon, layered fog drifting through pine-covered cliffs. Audio: thin alpine wind, distant raven caw echoing across stone, faint rustle of nylon parka, hushed orchestral swell as the camera rises.", "tags": [ "cinematic", "landscape", "crane-shot", "sunrise", "audio", "veo-official" ], "params": { "aspect": "21:9", "duration": "8s", "resolution": "4K" }, "source": { "name": "Google Cloud Blog — Ultimate Prompting Guide for Veo 3.1", "url": "https://cloud.google.com/blog/products/ai-machine-learning/ultimate-prompting-guide-for-veo-3-1", "tier": "official" }, "notes": "Google 官方「Crane Shot Canyon Hiker」原型。展示 crane shot 镜头语言 + 起伏构图。" }, { "id": "ve-007", "model": "veo-3", "category": "quiet-moments", "title": "雨夜巴士窗(浅景深情绪样板)", "prompt": "Moody cinematic style, cool blue tone palette, melancholic atmosphere. Subject: a young woman in her mid-twenties, dark hair tucked behind one ear, faint earphones in, gazing out a fogged bus window. Her reflection is faintly visible on the rain-streaked glass. Action: she traces a single line through the condensation, then lets her hand fall. Camera: tight close-up with very shallow depth of field, 85mm lens look, soft static framing. Lighting: cool blue ambient bath from passing streetlights, occasional warm amber from shopfronts smearing across the glass. Environment: interior of a city bus at night during a heavy rainstorm, blurred neon outside dissolving into bokeh. Audio: steady patter of rain on the bus roof, low diesel engine rumble, distant wet tire hiss, soft indie piano theme bleeding from her earphones.", "tags": [ "cinematic", "moody", "shallow-dof", "rain", "night", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud Blog — Ultimate Prompting Guide for Veo 3.1", "url": "https://cloud.google.com/blog/products/ai-machine-learning/ultimate-prompting-guide-for-veo-3-1", "tier": "official" }, "notes": "Google 官方「Bus Window Melancholic」原型,演示 shallow DOF + cool blue 情绪样板。" }, { "id": "ve-008", "model": "veo-3", "category": "dialogue-driven", "title": "黑色电影侦探办公室(双角色对白)", "prompt": "Film noir cinematic style, high contrast shadows, smoky atmosphere. Subject: a weathered detective in a rumpled trench coat sits behind a cluttered wooden desk, a mysterious woman in a long black coat stands in the doorway, half in silhouette. Action: detective looks up wearily from a stack of case files as the woman steps into the light. Camera: medium two-shot, slight low angle on the detective, framed through window-blind shadows. Lighting: single warm tungsten desk lamp, hard venetian-blind slats striping their faces, deep ink-black shadows. Environment: 1940s private investigator's office, ceiling fan turning slowly, rain-streaked window behind. Audio: distant city rain, ticking wall clock, the slow turn of the ceiling fan, faint mournful saxophone bleeding through the wall. Dialogue: Detective (weary): \"Of all the offices in this town, you had to walk into mine.\" Woman (slight mysterious smile): \"You were highly recommended.\"", "tags": [ "dialogue", "noir", "two-character", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud Blog — Ultimate Prompting Guide for Veo 3.1", "url": "https://cloud.google.com/blog/products/ai-machine-learning/ultimate-prompting-guide-for-veo-3-1", "tier": "official" }, "notes": "Google 官方 ingredients-to-video 双 shot 合并版,展示 Veo 多角色对白 + 情绪表演。" }, { "id": "ve-009", "model": "veo-3", "category": "dialogue-driven", "title": "舞台到观众 180° 弧形转场(带演唱)", "prompt": "Photorealistic cinematic concert style, dramatic stage lighting. Subject: a female pop star with shoulder-length dark hair and a sequined jacket, singing into a vintage chrome microphone. Action: she sings with eyes closed, then opens them as the camera arcs around her, ending with her view facing a sea of fans. Camera: smooth 180-degree arc shot starting on the front-facing close-up and circling around to a POV shot from behind her shoulders. Lighting: single dramatic front spotlight transitioning to bright stage lights with lens flare facing the crowd. Environment: dark concert stage with a roaring audience of silhouettes and phone lights below. Audio: live drum kit pulse, electric guitar sustain, crowd cheering and chanting rising in intensity, microphone reverb tail. Dialogue: she sings \"when you look me in the eyes, I can see a million stars.\"", "tags": [ "dialogue", "music", "performance", "arc-shot", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud Blog — Ultimate Prompting Guide for Veo 3.1", "url": "https://cloud.google.com/blog/products/ai-machine-learning/ultimate-prompting-guide-for-veo-3-1", "tier": "official" }, "notes": "Google 官方 first-and-last-frame workflow 转场示例,展示 180° arc 镜头与歌词同步生成。" }, { "id": "ve-010", "model": "veo-3", "category": "cinematic", "title": "丛林探险者多镜头时间码序列(8s 完整短片)", "prompt": "Cinematic adventure style, warm jungle palette, sense of discovery and wonder. Multi-shot sequence with timestamp prompting: [00:00-00:02] Medium shot from behind a young female explorer with a leather satchel and messy brown hair in a ponytail, as she pushes aside a large jungle vine to reveal a hidden path. [00:02-00:04] Reverse shot of the explorer's freckled face, her expression filled with awe as she gazes upon ancient, moss-covered ruins in the background. [00:04-00:06] Tracking shot following the explorer as she steps into the clearing and runs her hand over the intricate carvings on a crumbling stone wall. Emotion: wonder and reverence. [00:06-00:08] Wide, high-angle crane shot, revealing the lone explorer standing small in the center of the vast, forgotten temple complex, half-swallowed by the jungle. Lighting: dappled tropical sunlight breaking through canopy. Audio: the rustle of dense leaves and distant exotic bird calls; in the final beat a swelling, gentle orchestral score begins to play.", "tags": [ "multi-shot", "timestamp-prompting", "adventure", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud Blog — Ultimate Prompting Guide for Veo 3.1", "url": "https://cloud.google.com/blog/products/ai-machine-learning/ultimate-prompting-guide-for-veo-3-1", "tier": "official" }, "notes": "Google 官方 timestamp prompting 完整示例,Veo 3.1 独有能力:一次生成 4 个镜头组成的 8s 短片。" }, { "id": "ve-011", "model": "veo-3", "category": "portrait", "title": "生物发光萨满肖像(角色一致性)", "prompt": "Hyper-realistic, cinematic portrait style. Subject: a wise, androgynous shaman of indeterminate age. Their weathered skin is etched with intricate, bioluminescent circuit-like tattoos that pulse with a soft cyan light, deep set eyes, long silver braided hair adorned with small bone beads, draped in dark layered ceremonial robes. Action: the shaman opens their eyes slowly and inhales, the tattoos brightening with the breath. Camera: tight close-up portrait, very shallow depth of field, slow imperceptible push-in. Lighting: dominant key from the cyan tattoo glow on the face, deep negative-fill shadow on the opposite side, faint rim from above. Environment: pitch-dark sacred chamber, faint ember sparks floating in the air. Audio: low resonant hum from the tattoos, faint distant ritual chant, slow controlled breathing, occasional ember crackle. Dialogue: (in a calm whisper) \"The threads are listening tonight.\"", "tags": [ "portrait", "character", "cinematic", "audio", "dialogue", "veo-official" ], "params": { "aspect": "9:16", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud — Veo Prompt Guide (Vertex AI docs)", "url": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide", "tier": "official" }, "notes": "Vertex AI 官方 Veo prompt guide「Shaman Portrait」原型,演示 hyper-realistic + 角色一致性细节。" }, { "id": "ve-012", "model": "veo-3", "category": "action", "title": "古籍藏匿数据芯片(手部特写动作)", "prompt": "Cinematic thriller style, tight and tense atmosphere. Subject: a pair of black-gloved hands working over an ancient, leather-bound book; a small metallic data chip the size of a fingernail. Action: the gloved hand carefully slices open the spine of the book with a scalpel; the hand then delicately extracts the tiny metallic data chip hidden within the binding; the body tenses, listening. Camera: extreme close-up macro on the hands and scalpel, very shallow depth of field, slow drift-in. Lighting: single warm desk-lamp pool from frame-right, deep shadows around the edges. Environment: dimly lit private study, dust motes drifting, blurred shelves of leather books in the background. Audio: the soft crisp slice of the scalpel through old leather, dry paper crinkle, the metallic click of the chip lifting free, distant clock ticking, a sudden creak of a floorboard down the hall causing the hands to freeze.", "tags": [ "action", "macro", "thriller", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud — Veo Prompt Guide (Vertex AI docs)", "url": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide", "tier": "official" }, "notes": "Vertex AI 官方「Book Heist Scene」原型,演示动作连续性 + 紧张声音设计。" }, { "id": "ve-013", "model": "veo-3", "category": "experimental", "title": "生物发光遗忘之城(世界观氛围)", "prompt": "Cinematic dark-fantasy style, eerie and otherworldly atmosphere, painterly color palette. Subject: a rain-slicked, crumbling cobblestone street in a forgotten city shrouded in perpetual twilight, giant bioluminescent mushrooms of teal and magenta sprouting from cracked asphalt and broken balconies. Action: a slow forward dolly down the empty street, gentle drip of water off mushroom caps, a single glowing spore drifts past camera. Camera: low wide tracking shot at puddle level, slight Dutch tilt, anamorphic widescreen feel. Lighting: pulsating bioluminescent glow as the main key, faint moonlight diffused through fog as fill, mushroom caps casting soft colored pools on wet stone. Environment: derelict gothic buildings choked with vines, broken neon signs in an unknown script, distant towering ruins. Audio: dripping water, low ambient drone, faint distant wind chime, the soft pulse hum of the mushrooms breathing in sync with their glow.", "tags": [ "experimental", "dark-fantasy", "atmosphere", "world-building", "audio", "veo-official" ], "params": { "aspect": "21:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud — Veo Prompt Guide (Vertex AI docs)", "url": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide", "tier": "official" }, "notes": "Vertex AI 官方「Bioluminescent City」原型,展示 world-building + 同步声/光节拍。" }, { "id": "ve-014", "model": "veo-3", "category": "cinematic", "title": "古老罗盘戏剧性推镜(鼓点 + 烛光)", "prompt": "Cinematic mystery style, warm candlelit palette, slow dramatic pacing. Subject: a tarnished antique brass compass with a cracked glass face, lying on a dusty parchment map covered in faded ink markings, a single flickering candle beside it. Action: the compass needle twitches, then begins to spin slowly. Camera: slow dramatic zoom-in starting wide showing the map, candle, and weathered wooden desk, then smoothly pushing into the spinning needle until the brass casing fills the frame. Lighting: warm flickering candlelight as the only source, hard amber falloff into deep blacks, soft caustics on the brass. Environment: an explorer's study at night, leather-bound books and a quill in a brass inkwell just out of focus. Audio: the soft tick of the compass needle against glass, candle flame flutter, distant rolling thunder, low ominous string drone rising as the camera pushes in.", "tags": [ "cinematic", "mystery", "zoom-in", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud — Veo Prompt Guide (Vertex AI docs)", "url": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide", "tier": "official" }, "notes": "Vertex AI 官方「Mysterious Compass Zoom」原型,演示 dramatic zoom + 烛光主光。" }, { "id": "ve-015", "model": "veo-3", "category": "dialogue-driven", "title": "审讯室对峙(拉焦 + 双人对白)", "prompt": "Cinematic crime-drama style, harsh high-contrast lighting, sweat-and-grain texture. Subject: a seasoned detective in his fifties with stubble and a rumpled grey shirt, opposite a nervous informant in his thirties with sweat beading on his forehead. Action: detective leans forward across a metal table; informant flinches, looks down, then up. Camera: medium shot, then a slow rack focus shift from the detective in the foreground to the informant in the background. Lighting: single bare overhead bulb swinging slightly, hot specular on foreheads, deep black surrounds, faint blue spill from a one-way mirror. Environment: a dimly lit interrogation room, chipped concrete walls, a cup of cold coffee on the table. Audio: the slow swing creak of the bulb, the buzz of the filament, distant muffled radio chatter from the hallway, the informant's shallow breathing. Dialogue: Detective: \"Your story has holes.\" Informant: \"I'm telling you everything I know.\"", "tags": [ "dialogue", "crime-drama", "rack-focus", "two-character", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud — Veo Prompt Guide (Vertex AI docs)", "url": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide", "tier": "official" }, "notes": "Vertex AI 官方「Interrogation Room Dialogue」原型,演示 rack focus + 双人原生对白。" }, { "id": "ve-016", "model": "veo-3", "category": "nature", "title": "红玫瑰雨滴极特写(感官细节)", "prompt": "Hyper-realistic macro cinematic style, jewel-like color saturation. Subject: a single glistening drop of rain mid-fall, and a vibrant crimson red rose with velvet petals beaded with smaller droplets. Action: the drop strikes the upper petal, the petal trembles slightly and bends, sending two tiny secondary droplets skittering down the petal edge. Camera: extreme close-up macro, static lock-off, ultra shallow depth of field with only the impact petal in razor focus. Lighting: soft overcast diffused key from camera-left, faint cool rim from behind to catch the droplet's silver edge. Environment: a quiet garden after a passing shower, soft green bokeh of foliage in the background. Audio: an isolated crisp water-drop impact, a delicate petal rustle, faint distant rainfall, soft birdsong in the distance, gentle wind through leaves.", "tags": [ "nature", "macro", "slow-motion", "sensory", "audio", "veo-official" ], "params": { "aspect": "1:1", "duration": "6s", "resolution": "4K" }, "source": { "name": "Google Cloud — Veo Prompt Guide (Vertex AI docs)", "url": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide", "tier": "official" }, "notes": "Vertex AI 官方「Rain Drop Close-Up」原型,演示 macro + 感官音效细节。" }, { "id": "ve-017", "model": "veo-3", "category": "nature", "title": "玫瑰花苞绽放延时(静态长镜延时)", "prompt": "Cinematic botanical time-lapse style, painterly soft palette. Subject: a single red rose bud, its petals tightly closed at the start. Action: the flower slowly and gracefully unfurls over the course of the shot, petals rotating open in stages to reveal the vibrant inner layers, the stem subtly straightening toward the light. Camera: locked-off static close-up, no movement, perfectly centered on the bud. Lighting: warm soft directional sunlight slowly drifting from camera-left to camera-right across the frame, suggesting hours passing. Environment: studio black backdrop with a faint warm gradient, isolating the rose. Audio: a slow sustained ambient pad, subtle crackling and creaking of petal fibers stretching open, faint distant garden hum, a gentle final exhale as the flower reaches full bloom.", "tags": [ "nature", "timelapse", "botanical", "static-camera", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud — Veo Prompt Guide (Vertex AI docs)", "url": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide", "tier": "official" }, "notes": "Vertex AI 官方「Rose Unfurling Time-Lapse」原型,演示 temporal evolution + 静态相机。" }, { "id": "ve-018", "model": "veo-3", "category": "cinematic", "title": "城市天际线日转夜延时(建筑节奏)", "prompt": "Cinematic urban time-lapse style, rich color contrast between warm day and cool night. Subject: a bustling modern city skyline with glass skyscrapers, a winding elevated highway in the foreground. Action: a time-lapse of the city as day transitions to night — the sun sets casting long golden shadows across rooftops, then the city lights begin to twinkle on window by window, streams of car headlights forming red and white light trails along the highway. Camera: locked-off static wide shot, no movement. Lighting: from harsh midday daylight, through warm golden hour, into deep blue twilight, then sodium-orange streetlamp blanket. Environment: a metropolis viewed from a distant elevated rooftop, scattered clouds streaking across the sky. Audio: a sweeping cinematic orchestral score rising slowly, layered with distant city ambient hum, faint sirens passing far below, a soft wind across the rooftop.", "tags": [ "cinematic", "timelapse", "city", "day-to-night", "audio", "veo-official" ], "params": { "aspect": "21:9", "duration": "8s", "resolution": "4K" }, "source": { "name": "Google Cloud — Veo Prompt Guide (Vertex AI docs)", "url": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide", "tier": "official" }, "notes": "Vertex AI 官方「City Night Time-Lapse」原型,演示 day-to-night 时序 + 静态延时。" }, { "id": "ve-019", "model": "veo-3", "category": "experimental", "title": "动漫森林少女(明确风格指定)", "prompt": "Vibrant Japanese anime style, sharp clean line art, bright saturated colors, expressive cel-shaded faces, painterly background reminiscent of Studio Ghibli forests. Subject: a magical girl in her mid-teens with long flowing silver hair, glowing electric-blue eyes, a white and pale-blue ceremonial outfit with flowing ribbons, a tiny floating spirit creature drifting beside her shoulder. Action: she walks slowly through the forest, the wind lifting her hair, the spirit hopping playfully around her head; she pauses and reaches out to a glowing flower. Camera: smooth medium tracking shot from her side, slight parallax on the foreground ferns. Lighting: warm dappled sun shafts breaking through the canopy, soft volumetric god-rays catching airborne spores. Environment: a lush enchanted forest with oversized glowing flowers, ancient mossy stone lanterns, drifting pollen sparkles. Audio: a whimsical orchestral theme with woodwind and harp, soft footsteps on moss, the tinkling chime of the spirit's movement, a gentle breeze through leaves, the magical girl humming a soft melody.", "tags": [ "experimental", "anime", "stylized", "tracking-shot", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud — Veo Prompt Guide (Vertex AI docs)", "url": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/video-gen-prompt-guide", "tier": "official" }, "notes": "Vertex AI 官方「Anime Forest Scene」原型,演示明确风格锚定 + 角色细节描述。" }, { "id": "ve-020", "model": "veo-3", "category": "dialogue-driven", "title": "巴黎档案馆历史学家(角色一致性 + 单人独白)", "prompt": "Cinematic mystery drama style, dimly lit warm palette, classic European film grade. Subject: Clara, a historian in her early 30s, with observant dark brown eyes that hold a quiet intensity, chin-length black hair styled in a classic bob, dressed in a sophisticated dark navy-blue wool coat with a silk scarf patterned with subtle gold and cream designs tied around her neck. Action: she stands before a large ancient wooden table and carefully turns the fragile yellowed page of a massive leather-bound book, her expression one of deep concentration. Camera: medium shot, slow dolly-in. Lighting: a single beam of light from a high window casting visible dust motes, deep mahogany shadows surrounding her. Environment: a dimly lit grand Parisian archive, towering shelves of leather books, a brass desk lamp glowing faintly. Audio: the soft crinkle of ancient parchment, distant church bells outside, the slow tick of an antique pendulum clock, a low ambient cello drone. Dialogue: Clara, in a voice that is crisp and clear with a thoughtful analytical tone and a standard American accent, says: \"It has to be here.\"", "tags": [ "dialogue", "character-consistency", "mystery", "dolly-in", "audio", "veo-official" ], "params": { "aspect": "16:9", "duration": "8s", "resolution": "1080p" }, "source": { "name": "Google Cloud — Veo Best Practices (Vertex AI docs)", "url": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/best-practice", "tier": "official" }, "notes": "Vertex AI 官方 best-practice「Clara / Parisian Archive」原型,展示角色一致性完整描述公式(姓名+年龄+体貌+服饰+声线+口音)以便跨镜头复用。" }, { "id": "wn-001", "model": "wan-2.5", "category": "lip-sync", "title": "Wan 数字人对口型样板", "prompt": "Entity: A young Asian female presenter in a sleek navy blazer, hair tied back neatly.\nScene: Modern minimalist office with soft white background, warm rim light.\nMotion: She faces the camera directly, lips synced to dialogue, occasional natural hand gestures, slight smile.\nSound: Voice: \"Welcome back to our weekly product update. This week we shipped three major features.\" Confident professional female voice, room ambient hum, no music.", "tags": [ "digital-human", "lip-sync", "presenter", "wan-official" ], "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Alibaba Cloud", "url": "https://www.alibabacloud.com/help/en/model-studio/text-to-video-prompt", "tier": "official" }, "notes": "Wan 2.7 看家本领:数字人 lip-sync,业界对口型最准" }, { "id": "wn-002", "model": "wan-2.5", "category": "dialogue-driven", "title": "Wan 公式样板(Entity+Scene+Motion+Sound)", "prompt": "Entity: A male barista with a short beard and apron.\nScene: A bustling third-wave coffee shop, morning light through large windows.\nMotion: He steams milk with practiced motion, taps the pitcher, pours latte art into a white ceramic cup, looks up and slides it across the counter.\nSound: Steam wand hissing, milk frothing, cup clinking, soft indie pop background music, brief customer chatter.", "tags": [ "product", "audio", "wan-official", "formula" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Alibaba Cloud", "url": "https://www.alibabacloud.com/blog/model-studio-wan-video-generation-prompts-recipe_602777", "tier": "official" } }, { "id": "wn-003", "model": "wan-2.5", "category": "music-video", "title": "Wan 音乐 MV 同步节拍", "prompt": "Entity: A solo female singer in a flowing white dress.\nScene: Empty old theater stage with single spotlight, dust particles in the beam.\nMotion: She lifts the microphone, sings emotionally, swaying gently with the music, hair catches the spotlight.\nSound: A solo piano ballad with subtle string section, her clear vocal singing 'In the silence I find my voice', perfectly synced to mouth movement, slight reverb.", "tags": [ "music-video", "audio", "lip-sync", "emotional" ], "params": { "aspect": "9:16", "duration": "10s" }, "source": { "name": "Kie.ai Wan", "url": "https://kie.ai/wan-2-5", "tier": "reputable-3rd-party" } }, { "id": "wn-004", "model": "wan-2.5", "category": "product-commercial", "title": "Wan 短广告同步配音", "prompt": "Entity: A premium black skincare serum bottle on a marble pedestal.\nScene: Soft white studio with gentle blue rim light.\nMotion: Camera slowly orbits the bottle 360 degrees, then pushes in to a macro close-up on the gold logo.\nSound: Warm female voiceover: 'Discover the science of timeless skin.' Subtle ambient piano underscore, faint glass clink at end.", "tags": [ "product", "voiceover", "audio", "luxury" ], "params": { "aspect": "1:1", "duration": "8s" }, "source": { "name": "Wan", "url": "https://kie.ai/wan-2-5", "tier": "reputable-3rd-party" } }, { "id": "wn-005", "model": "wan-2.5", "title": "Wan 双人对话(情侣晚餐)", "prompt": "Entity: A man and a woman in their 30s, both in casual but elegant attire, seated across a candlelit dining table.\nScene: A vintage wood-paneled dining room with dark wooden walls and a large window revealing lush green plants outside. Sunlight streams through the window, casting gentle side lighting across the scene. The table holds bread, grapes, silver cutlery, and a gleaming tea pot. Warm tungsten ambient glow blends with the daylight.\nMotion: She lifts her wine glass slightly and looks toward him with a faint, sad smile. He exhales softly, looks down, then back up. Lips synced to dialogue. Medium two-shot, camera locked off.\nSound: The man says in a low, restrained tone: 'We can't keep pretending everything is unchanged.' The woman exhales softly, then murmurs: 'But what if forgetting hurts more than remembering?' Warm room tone, faint birdsong outside the window, no music.", "tags": [ "dialogue", "romantic", "audio", "two-shot" ], "category": "dialogue-driven", "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Alibaba Cloud Wan Prompts Recipe", "url": "https://www.alibabacloud.com/blog/model-studio-wan-video-generation-prompts-recipe_602777", "tier": "official" } }, { "id": "wn-006", "model": "wan-2.5", "category": "lip-sync", "title": "Wan 新闻播报数字人", "prompt": "Entity: A male news anchor in a tailored grey suit and blue tie.\nScene: Modern broadcast studio with curved LED background showing world map graphics.\nMotion: He looks directly at camera, formal posture, lips synced precisely to dialogue, slight gesture mid-sentence.\nSound: Professional male anchor voice: 'Good evening. Tonight's top stories include the global tech summit and...' Studio ambient hum, no music.", "tags": [ "digital-human", "lip-sync", "news", "professional" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Wan", "url": "https://www.alibabacloud.com/help/en/model-studio/use-video-generation", "tier": "official" } }, { "id": "wn-007", "model": "wan-2.5", "title": "Wan I2V 风景动起来(带音频)", "prompt": "Entity: The autumn forest in the reference image, rich oranges and yellows against a backdrop of slender tree trunks.\nScene: A quiet stretch of woodland in late afternoon, low golden sunlight filtering between the trees and casting long shadows across the leaf-strewn ground.\nMotion: The camera pushes slowly forward through the trees on a steady glide. Individual leaves drift down gently, each one rotating and turning as it falls. Beams of sunlight shift subtly with the breeze, mottled patches of light moving across the forest floor.\nSound: Crisp rustling of dry leaves underfoot, light wind through branches, the distant call of a single songbird, faint creak of swaying timber. No music.", "tags": [ "image-to-video", "nature", "audio", "atmospheric" ], "category": "image-to-video", "params": { "aspect": "16:9", "duration": "8s" }, "source": { "name": "Alibaba Cloud Wan Prompts Recipe", "url": "https://www.alibabacloud.com/blog/model-studio-wan-video-generation-prompts-recipe_602777", "tier": "official" } }, { "id": "wn-008", "model": "wan-2.5", "title": "Wan TikTok 自拍解说", "prompt": "Entity: A young female content creator in her early 20s, wearing a casual hoodie, holding her phone selfie-style facing the camera directly.\nScene: A trendy minimalist bedroom with white walls and a single neon LED strip running along the ceiling, casting a soft purple-pink wash. Vertical 9:16 framing.\nMotion: She talks directly to the phone camera with energetic gestures — pointing, shrugging, raising both hands. Her facial expressions shift with the cadence of her speech. Lips synced to dialogue. Subtle handheld bob from holding the phone.\nSound: She says in a bright, conversational tone: 'Okay, so the trick I've been using all week is finally going to be in this video — stick around for the last part.' Quick, upbeat cadence, slight reverb of a small bedroom, no background music.", "tags": [ "tiktok", "lip-sync", "creator", "vertical" ], "category": "social-viral", "params": { "aspect": "9:16", "duration": "8s" }, "source": { "name": "Alibaba Cloud Model Studio (Wan)", "url": "https://www.alibabacloud.com/help/en/model-studio/text-to-video-prompt", "tier": "official" } }, { "id": "wn-009", "model": "wan-2.5", "category": "comedy", "title": "Wan 卡通角色配音", "prompt": "Entity: A 3D cartoon corgi character with oversized expressive eyes.\nScene: A cozy stylized kitchen, soft pastel colors.\nMotion: Corgi stands on hind legs, opens fridge, looks shocked, lips animated to dialogue.\nSound: Squeaky high-pitched corgi voice: 'There's no more bacon? This is a tragedy!' Cartoon piano sting, fridge door creak, faint snare drum punctuation.", "tags": [ "cartoon", "comedy", "audio", "lip-sync" ], "params": { "aspect": "16:9", "duration": "6s" }, "source": { "name": "Wan", "url": "https://kie.ai/wan-2-5", "tier": "reputable-3rd-party" } }, { "id": "wn-010", "model": "wan-2.5", "category": "documentary", "title": "Wan 真实采访风(带环境音)", "prompt": "Entity: An elderly fisherman with weathered face, wearing a yellow rain slicker.\nScene: Misty harbor at dawn, fishing boats moored behind him.\nMotion: He speaks calmly to the camera, lips synced naturally.\nSound: His thoughtful voice with slight accent: 'I've been at sea since I was twelve. The ocean teaches you everything.' Boat ropes creaking, distant gulls, gentle waves against hull, no music.", "tags": [ "documentary", "interview", "audio", "atmospheric" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Wan", "url": "https://blog.segmind.com/wan-i2v-prompts-tips-guide/", "tier": "reputable-3rd-party" } }, { "id": "wn-011", "model": "wan-2.5", "title": "Wan 体育解说(带音效)", "prompt": "Entity: A football player mid-stride about to take a free kick, jersey number visible, focused expression.\nScene: A floodlit stadium at night, packed stands behind him glowing with phone lights and team scarves, pristine green pitch underfoot.\nMotion: He plants his foot, body twists, strikes the ball with power. The ball curves through the air toward goal, follow-through compresses his standing knee with accurate weight. Camera tracks the ball from low behind him.\nSound: Crisp impact of boot on ball, immediate roar of the crowd swelling in pitch, an announcer's voice in mid-tempo English: 'And he steps up — that's curling, curling toward goal!' Ambient stadium chant just below the commentary, no music.", "tags": [ "sports", "audio", "commentary", "stadium" ], "category": "action", "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Alibaba Cloud Wan Prompts Recipe", "url": "https://www.alibabacloud.com/blog/model-studio-wan-video-generation-prompts-recipe_602777", "tier": "official" } }, { "id": "wn-012", "model": "wan-2.5", "category": "chinese-native", "title": "Wan 中文场景:茶室对话", "prompt": "Entity: 两位年长茶师,身穿素色亚麻茶服。\nScene: 中式茶室,红木茶台,窗外竹影摇曳。\nMotion: 一位茶师注水入茶壶,另一位欣赏茶汤色泽,两人轻声交流。\nSound: 茶师 A 温和说道:'今年的明前龙井,香气格外清雅。' 茶师 B 点头:'是啊,雨水恰到好处。' 注水声、茶杯轻碰声、远处古琴隐约。", "tags": [ "chinese", "tea", "dialogue", "audio" ], "params": { "aspect": "16:9", "duration": "10s" }, "source": { "name": "Wan", "url": "https://www.alibabacloud.com/help/en/model-studio/text-to-video-prompt", "tier": "official" } }, { "id": "k4-001", "model": "kling-4.0", "category": "technique-snippet", "title": "Kling 4.0 四世界罗盘 Match Cut", "prompt": "[OUTPUT]\n\n12 seconds, 16:9, four-scene cinematic match-cut montage, crisp realistic materials, exactly four shots.\n\n[REFERENCE ROLES]\n\nObject reference locks one palm-sized scratched copper compass with a dark blue needle and plain lid. Environment references control only: coastal cliff at dawn, night train table, desert survey camp, snowy observatory. Do not transfer people or objects between environment references.\n\n[MATCH-CUT RULE]\n\nThe compass center remains at 52% frame width and 58% frame height, same apparent size and same 35-degree lid angle at every cut. Each transition occurs during the final six frames of a hand closing the lid; motion direction stays right-to-left.\n\n[TIMED SHOTS]\n\n0–3s — Coastal cliff macro. Weathered adult hand opens the compass; needle settles north, salt mist crosses left-to-right, slow 5cm push-in.\n\n3–6s — Cut under closing fingers to the same compass on a night-train table. Different sleeve, same hand path. Passing lights sweep across copper; needle trembles once with rail vibration.\n\n6–9s — Cut under the lid to a desert survey cloth. Fine sand moves around, never onto, the hinge. Hand rotates the whole compass 20 degrees clockwise; needle independently returns north.\n\n9–12s — Cut under closing fingers to snowy observatory stone. Gloved hand opens the lid fully; camera tilts from compass toward a clear polar star, ending with the compass still visible at lower center.\n\n[AUDIO]\n\nOne continuous low musical note bridges all cuts. Distinct layers: surf and lid click; rail rhythm; dry wind; high mountain air. The same metal click lands exactly at 2.8s, 5.8s and 8.8s.\n\n[CONSTRAINTS]\n\nOne compass only. Preserve scratches, hinge, needle color, scale and screen position. Cuts happen only behind fingers and lid; no dissolve, portal, morphing map, readable text, logo or watermark.", "tags": [ "kling-4.0", "technique-snippet", "match-cut", "reference", "night", "rain" ], "params": { "aspect": "16:9", "duration": "12s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/reference-editing-and-control.md", "tier": "reputable-3rd-party" }, "notes": "omni reference 范式:一个物体 reference 加四个环境 reference 各司其职,match-cut 全部藏进合盖动作,是 4.0 多引用叙事的标杆写法。" }, { "id": "k4-002", "model": "kling-4.0", "category": "quiet-moments", "title": "Kling 4.0 三代人午餐", "prompt": "[OUTPUT]\n\n15 seconds, 16:9, intimate family lunch scene, naturalistic cinema, three connected shots with native dialogue.\n\n[REFERENCE ROLES]\n\nImage 1 locks Mei, adult grandmother with short white curls and burgundy cardigan. Image 2 locks adult daughter An, chin-length black hair and olive shirt. Image 3 locks adult grandson Leo, curly dark hair and mustard sweater. Image 4 supplies only the small sunlit dining room and oval wooden table; ignore any people in it.\n\n[GEOGRAPHY]\n\nMei sits frame left, An frame right, Leo centered on far side. One blue serving bowl in the table center. Their seats, wardrobe and eyelines never swap.\n\n[TIMED SHOTS]\n\n0–5s — Wide locked master. Mei lifts the blue bowl with both hands and passes it clockwise to An. Leo watches the bowl, then glances at Mei.\n\n5–10s — Medium two-shot across table. An receives the bowl from below, sets it down once and says in Mandarin: “还是这个味道。” Mei is soft foreground left, listening without speaking.\n\n10–15s — Close on Mei. She smiles with restrained pride, looks toward An and answers in English: “You remembered.” Focus shifts to Leo quietly taking one dumpling from the bowl; end on all three sharing a small laugh.\n\n[AUDIO]\n\nQuiet midday room, ceramic contact, chair fabric and distant neighborhood birds. Only An speaks the Mandarin line; only Mei speaks the English line. No narrator or subtitles.\n\n[CONSTRAINTS]\n\nExactly three adults, one bowl and one dumpling lifted. Natural hand contact and table occlusion. Preserve identity, age, hair, clothing, seat and screen direction. No face blending, duplicate people, extra dishes appearing, generated wall text, logo or watermark.", "tags": [ "kling-4.0", "quiet-moments", "dialogue", "reference", "rain" ], "params": { "aspect": "16:9", "duration": "15s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/reference-editing-and-control.md", "tier": "reputable-3rd-party" }, "notes": "三人物 reference 加一房间 reference 的身份锁定写法,中英双语原生对白,展示多角色一致性控制。" }, { "id": "k4-003", "model": "kling-4.0", "category": "cinematic", "title": "Kling 4.0 末班车归家", "prompt": "[OUTPUT]\n\n15 seconds, 16:9, three-shot contemporary drama, restrained naturalism, cool winter night with warm practical bus light.\n\n[CONTINUITY ANCHORS]\n\nEli: adult man in his early 30s, narrow face, short curly dark hair, charcoal wool coat, faded red scarf, tired eyes.\n\nJune: adult woman in her late 50s, silver bob, moss-green quilted jacket, canvas grocery bag, small brass house key in her right hand.\n\nKeep faces, ages, wardrobe, bag, key and left-to-right screen direction unchanged.\n\n[WORLD]\n\nNearly empty suburban bus at night. Rain tracks down the windows. Eli sits frame left near the rear door; June stands in the aisle frame right. Sodium streetlights sweep across the interior as the bus moves.\n\n[TIMED SHOTS]\n\n0–5s — Medium two-shot from the opposite seat, camera locked with gentle vehicle vibration. June notices the red scarf, stops one step away and tightens her hand around the key. Eli looks up slowly. Neither speaks.\n\n5–10s — Over June’s shoulder into Eli’s close-up; keep his eyeline toward frame right. The passing streetlight reveals recognition. He removes one earbud and says, quietly: “You still have the key.”\n\n10–15s — Reverse close-up on June, eyeline frame left. Her guarded expression softens; she opens her palm to show the brass key and answers: “You never asked for it back.” Hold one second after the line as the bus brakes gently and both sway in the same direction.\n\n[AUDIO]\n\nContinuous engine hum, soft rain on glass and two distant stop-request chimes. Eli (English, low and hesitant), then June (English, warm but controlled). No music until the final second, when one low cello note enters beneath the bus brake.\n\n[CONSTRAINTS]\n\nOnly the named speaker’s mouth moves. Preserve the red scarf and brass key. Natural blinking and breathing; no crying, embrace or exaggerated smile. No subtitles, signage text, logos, face drift, age change, jump in seat positions or sudden weather change.", "tags": [ "kling-4.0", "cinematic", "close-up", "night", "rain" ], "params": { "aspect": "16:9", "duration": "15s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/cinematic-and-dialogue.md", "tier": "reputable-3rd-party" }, "notes": "三镜头情感短片:道具(钥匙/围巾)锚定一致性,眼神线和环境声床贯穿全片,电影感叙事的完整模板。" }, { "id": "k4-004", "model": "kling-4.0", "category": "dialogue-driven", "title": "Kling 4.0 站台纸鹤(韩西双语)", "prompt": "[OUTPUT]\n\n15 seconds, 16:9, four-shot independent-film reunion, Korean and Spanish dialogue, native audio, subtle film grain.\n\n[CHARACTER MAP]\n\nMina: adult Korean woman, shoulder-length wet black hair, rust-red raincoat, small kraft-paper crane in both hands, low warm voice.\n\nÁlex: adult Spanish man, wavy dark hair, short beard, navy railway work jacket with one pale reflective stripe and no logo, soft tenor voice.\n\nThe paper crane has one blue ink dot on its left wing; preserve it in every shot.\n\n[WORLD]\n\nSmall covered coastal train platform at blue hour. Cream-and-teal local train is stationary frame right; rough sea frame left. Warm overhead lamps reflect on wet tiles. Rain blows diagonally beyond the roof.\n\n[TIMED SHOTS]\n\n0–4s — Wide symmetrical platform shot. Mina enters from frame left and stops; Álex steps down from the train at frame right. They recognize each other across six meters. Camera makes a very slow push forward.\n\n4–7s — Close-up on Mina, Álex remains a soft silhouette deep frame right. She unfolds her fingers to reveal the paper crane. Her inhale catches; she says in Korean, almost a whisper: “늦었네.”\n\n7–10s — Shot/reverse-shot close-up on Álex, eyeline frame left. He exhales, gives one small relieved nod and replies in Spanish: “Pero llegué.”\n\n10–15s — Medium-wide two-shot from the side. They each take one step closer but do not embrace. Mina extends the crane; Álex reaches out and stops just before touching it. Hold on the charged gap between their hands.\n\n[AUDIO]\n\nContinuous coastal rain, distant surf and low idling-train hum. No station announcement. Keep Mina’s Korean and Álex’s Spanish exactly in their written languages; no translation. A sparse two-note piano motif begins only after Álex finishes.\n\n[CONSTRAINTS]\n\nOne speaking mouth at a time. Preserve faces, coat colors, train position, rain direction and matched eyelines. The paper crane never changes shape, color or hand. Restrained adult emotion; no melodrama. No subtitles, readable signs, logos, extra passengers, facial drift, dry wardrobe between shots or camera-axis reversal.", "tags": [ "kling-4.0", "dialogue-driven", "dialogue", "close-up", "rain" ], "params": { "aspect": "16:9", "duration": "15s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/cinematic-and-dialogue.md", "tier": "reputable-3rd-party" }, "notes": "韩语加西班牙语双语对白、逐发话人绑定声线与台词,是 4.0 九语言 lip sync 能力的典型用法。" }, { "id": "k4-005", "model": "kling-4.0", "category": "product-commercial", "title": "Kling 4.0 植物气泡饮品广告", "prompt": "[OUTPUT]\n\n9 seconds, 16:9, premium non-alcoholic sparkling botanical drink commercial, photorealistic macro finish, three shots.\n\n[PRODUCT LOCK]\n\nOne smoked-green glass bottle with a narrow shoulder, silver knurled cap and blank embossed oval on the front. No printed label or text. Preserve bottle silhouette, cap, oval, glass color and fill level exactly across all shots.\n\n[WORLD]\n\nWet black basalt surface in a dark studio. Deep emerald rim light from frame left, cool silver key light overhead, narrow warm beam from frame right. Fine mist moves left to right.\n\n[TIMED SHOTS]\n\n0–3s — 100mm macro, low three-quarter angle. Camera slides slowly right while staying focused on condensation. One droplet forms at the shoulder and runs straight down; bottle remains motionless and vertical.\n\n3–6s — Match cut to overhead close-up. A single pale-green citrus peel enters from frame left and completes one smooth clockwise spiral around the bottle without touching it. Fine bubbles rise inside the liquid at constant speed.\n\n6–9s — Medium hero shot. Camera performs a slow 25-degree clockwise orbit at constant distance. The cap lifts two millimeters with a clean pressure release; a controlled halo of cold vapor blooms and clears. End with bottle centered and oval facing camera.\n\n[AUDIO]\n\n0–3s: low studio silence plus individual water drops. 3.1s: crisp peel swish. 6.4s: short cap click and delicate carbonation release. 8s: two warm glass-marimba notes, then silence.\n\n[CONSTRAINTS]\n\nThe bottle never floats, bends, rotates independently or changes scale. Keep contact shadow attached to basalt. Physically plausible refraction, droplets, vapor and peel trajectory. No splash covering the product, alcohol cues, glassware, people, text, logo, watermark, label mutation or duplicate bottle.", "tags": [ "kling-4.0", "product-commercial", "close-up", "rain", "product" ], "params": { "aspect": "16:9", "duration": "9s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/commercial-and-ugc.md", "tier": "reputable-3rd-party" }, "notes": "产品锁定写法:瓶身几何、冷凝水滴、柑橘皮各有独立运动指令,结尾开盖蒸汽有精确时间点,电商级广告模板。" }, { "id": "k4-006", "model": "kling-4.0", "category": "music-video", "title": "Kling 4.0 屋顶打击乐", "prompt": "[OUTPUT]\n\n15 seconds, 16:9, live rooftop percussion performance at dawn, documentary concert energy, original rhythm, four shots.\n\n[PERFORMER MAP]\n\nAya: adult Japanese woman, red windbreaker, short hair, frame left, plays one shallow metal handpan.\n\nMalik: adult British man, blue overshirt, shaved head, center, plays one wooden cajón.\n\nCamila: adult Colombian woman, yellow knit top, long braid, frame right, plays two small seed shakers.\n\nPreserve faces, instruments, clothing colors and left-center-right positions.\n\n[RHYTHM AND SHOTS]\n\n0–4s — Wide locked shot, pale sunrise behind city roofs. Malik starts a steady low cajón pulse: four evenly spaced hits. The others listen and sway naturally.\n\n4–8s — Medium side track from Aya to Malik. Aya adds a three-note handpan phrase after every second cajón hit. Her fingers contact visible tone fields; each note aligns exactly to contact.\n\n8–12s — Close-up on Camila’s hands and face. She adds a soft double-shake pattern, wrists moving compactly. Camera rack-focuses from seeds inside the translucent shaker to her smile.\n\n12–15s — Return to wide. All three play one synchronized final accent, then freeze their hands while the metal note decays. A flock of ordinary birds crosses far background after the final hit.\n\n[AUDIO]\n\nNative performance only: dry low cajón, warm resonant handpan, fine seed texture, rooftop wind and distant city. No sampled song, vocals, cheering or added orchestra. Maintain room-free outdoor acoustics and realistic decay.\n\n[CONSTRAINTS]\n\nOne instrument per performer; accurate hand-to-sound synchronization; fixed stage positions. No extra sticks, changing instruments, impossible hand speed, duplicated fingers, lip movement, visible brand, skyline text or watermark.", "tags": [ "kling-4.0", "music-video", "close-up", "rain" ], "params": { "aspect": "16:9", "duration": "15s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/style-and-performance.md", "tier": "reputable-3rd-party" }, "notes": "原生音频音乐表演:三乐器的节奏全部写进时间轴、每个音符对应可见的接触动作,声画同步的标杆写法。" }, { "id": "k4-007", "model": "kling-4.0", "category": "sports", "title": "Kling 4.0 湿地赛道精准圈速", "prompt": "[OUTPUT]\n\n15 seconds, 21:9, realistic closed-course driver-training film at dawn, four shots, high energy without collisions.\n\n[VEHICLE AND DRIVER]\n\nOne compact cobalt-blue track car with white roof, black seven-spoke wheels, blank number panel and no logo. Adult professional driver wears a white certified helmet, blue suit and secured six-point harness. Preserve vehicle panels, wheel design and driver throughout.\n\n[TRACK GEOGRAPHY]\n\nPrivate wet handling circuit, right-hand hairpin marked by six orange cones and a white apex line. Empty runoff, barriers and marshals safely behind fence. Car travels left-to-right before turning away from camera.\n\n[TIMED SHOTS]\n\n0–4s — Interior footwell insert. Right foot releases brake progressively and applies throttle; left foot stays on dead pedal. Water beads move backward on side window. Engine note rises smoothly.\n\n4–8s — Low exterior side tracking. Car approaches left-to-right at controlled speed; tire spray trails behind wheels, never ahead. Camera remains outside barriers.\n\n8–12s — High fixed hairpin view. Driver brakes before cone one, turns once toward the apex and follows the white line without sliding or hitting cones. Brake lights illuminate during deceleration only.\n\n12–15s — Rear three-quarter from safe long lens. Car straightens, accelerates away and spray narrows as it reaches drier pavement. End with all six cones standing.\n\n[AUDIO]\n\nAuthentic single engine layer matched across cuts, wet tire hiss, two gear changes and no music until a restrained pulse enters at 12s. No dialogue.\n\n[CONSTRAINTS]\n\nPrivate course, one car, trained adult, correct harness and intact safety zone. No public traffic, passenger, collision, drift, jump, hydroplane, cone strike, speed claim, dashboard text, badge or watermark.", "tags": [ "kling-4.0", "sports", "dialogue", "rain" ], "params": { "aspect": "21:9", "duration": "15s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/genre-action-and-sound.md", "tier": "reputable-3rd-party" }, "notes": "21:9 宽画幅赛车短片:赛道地理、喷水方向、刹车灯时机、锥桶数量全部量化,高速动作的稳定写法。" }, { "id": "k4-008", "model": "kling-4.0", "category": "social-viral", "title": "Kling 4.0 归来的地铁卡", "prompt": "[OUTPUT]\n\n15 seconds, 9:16, contemporary vertical micro-drama, restrained realism, four fast but readable shots, episode-one cliffhanger.\n\n[CONTINUITY ANCHORS]\n\nMina: adult woman, short black bob, camel coat, canvas tote on left shoulder. Theo: adult station attendant, shaved head, navy uniform with no badge. Prop: one scratched green transit card with a small crescent-shaped corner chip; no readable printing.\n\n[GEOGRAPHY]\n\nQuiet underground station after the evening rush. Mina stands outside the closed service window, Theo inside. The card begins in Mina's right hand. Camera never crosses the service-window axis.\n\n[TIMED SHOTS]\n\n0–4s — Tight handheld medium. Mina slides the green card through the tray and says in Mandarin: “这张卡不是我的。” Theo looks down, then freezes before touching it.\n\n4–8s — Insert from Mina's side. Theo turns the card over with one gloved fingertip. The crescent chip is clearly visible. He asks quietly in English: “Where did you find this?”\n\n8–12s — Close on Mina. Her defensive expression softens; she answers: “Inside my locked apartment.” A train passes behind camera, briefly sweeping light across her face without hiding the cut.\n\n12–15s — Close behind Theo's shoulder. He opens a shallow drawer containing one second, identical chipped green card in a clear evidence sleeve. He does not remove it. Mina sees it and steps back once. Cut to black before either speaks.\n\n[AUDIO]\n\nStation ventilation, distant train, tray scrape and drawer rail. Only Mina speaks the Mandarin line and final English line; only Theo asks the question. One low tonal hit when the drawer opens; no subtitles.\n\n[CONSTRAINTS]\n\nTwo adults and exactly two matching cards by the final shot. Preserve chip position, wardrobe, tote shoulder, window axis and eyelines. No police insignia, threat, weapon, supernatural glow, generated text, logo or watermark.", "tags": [ "kling-4.0", "social-viral", "rain" ], "params": { "aspect": "9:16", "duration": "15s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/short-drama-and-viral-social.md", "tier": "reputable-3rd-party" }, "notes": "竖屏 9:16 微剧悬念:缺口地铁卡道具贯穿四镜、中英对白、切黑收尾,短剧第一集病毒式模板。" }, { "id": "k4-009", "model": "kling-4.0", "category": "nature", "title": "Kling 4.0 茶杯里的风暴", "prompt": "[OUTPUT]\n\n8 seconds, 1:1, one locked macro shot, photorealistic tabletop surrealism, physically coherent miniature storm.\n\n[SCENE]\n\nPlain white porcelain teacup on a worn oak table beside a closed book. Tea surface visible from a high three-quarter angle. Warm afternoon window light from frame left; cup handle points right. No branding or text.\n\n[ACTION TIMELINE]\n\n0–2s — Tea is almost still. One tiny circular cloud forms two centimeters above the liquid, rotating clockwise. The room outside the cup remains calm.\n\n2–5s — Cloud darkens and releases localized rain only into the cup. Concentric ripples spread to the porcelain edge. Three miniature lightning filaments illuminate the tea from within, each separated by half a second; no bolt leaves the cup.\n\n5–7s — A thumb-sized waterspout rises from the center and rotates once, pulling three loose tea leaves into orbit without spilling liquid over the rim.\n\n7–8s — Storm collapses into one soft puff; surface settles toward stillness. A final single droplet falls into the exact center for a loopable end.\n\n[AUDIO]\n\nQuiet room, magnified rain taps, three delicate electrical ticks synchronized to light, low water swirl, final droplet. No thunder boom, dialogue or music.\n\n[CONSTRAINTS]\n\nCup, handle, book, table, camera and room light remain fixed. Weather is contained entirely within cup boundaries. Realistic reflections and ripple propagation. No overflow, broken cup, fire, giant ocean, camera movement, text, logo or watermark.", "tags": [ "kling-4.0", "nature", "dialogue", "rain" ], "params": { "aspect": "1:1", "duration": "8s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/action-and-vfx.md", "tier": "reputable-3rd-party" }, "notes": "微观天气奇观:风暴能量严格限定在杯内并按时间衰减,结尾可循环,尺度反差的 VFX 写法。" }, { "id": "k4-010", "model": "kling-4.0", "category": "documentary", "title": "Kling 4.0 极光之上的静默对接", "prompt": "[OUTPUT]\n\n15 seconds, 21:9, realistic near-future orbital docking visualization, calm procedural tone, three shots.\n\n[VEHICLE LOCKS]\n\nService craft: small white cylinder, four folded rectangular solar panels, black circular docking ring at nose, no crew visible. Station: gray research module with matching axial port and two long blue solar arrays. Blank surfaces, no national or company marks.\n\n[ORIENTATION]\n\nStation docking axis runs screen left-to-right. Station port faces frame left; service craft approaches from frame left nose-first. Earth limb remains low background with a faint auroral band. Sunlight direction stays upper right.\n\n[TIMED SHOTS]\n\n0–5s — Long-lens side view. Craft closes distance slowly at less than one body length over five seconds. Two tiny lateral thruster pulses correct upward drift; exhaust is brief and subtle. Station remains inertially steady.\n\n5–10s — Camera cuts to station-port view looking outward. Craft nose stays centered using one short roll correction under five degrees, then stops all visible rotation. No rapid zoom.\n\n10–15s — Exterior three-quarter. Docking rings meet at near-zero relative speed, compress slightly and lock with one small mechanical movement. Solar arrays do not flex. End with both vehicles moving as one while Earth drifts slowly behind.\n\n[AUDIO]\n\nExterior shots have no diegetic sound. A restrained non-diegetic tonal bed continues. At contact, cut briefly to an interior structural microphone perspective for one muted latch thump and two confirmation clicks; no voice-over.\n\n[CONSTRAINTS]\n\nVisualization, not operational instruction. Preserve ports, axes, solar panels, light and low relative speed. No explosion, debris, visible exterior sound wave, large flame, astronaut, flag, readable UI, mission claim, logo or watermark.", "tags": [ "kling-4.0", "documentary", "rain", "explosion" ], "params": { "aspect": "21:9", "duration": "15s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/genre-action-and-sound.md", "tier": "reputable-3rd-party" }, "notes": "轨道对接物理可视化:固定轴线、低相对速度、接触瞬间切内部麦克风视角收音,硬科幻纪录片模板。" }, { "id": "k4-011", "model": "kling-4.0", "category": "fantasy-scifi", "title": "Kling 4.0 玻璃蝠鲼盐峡追逐", "prompt": "[OUTPUT]\n\n12 seconds, ultra-wide science-fantasy pursuit, photorealistic materials, three shots, fast but readable action, no weapons.\n\n[CONTINUITY ANCHORS]\n\nCourier Nara: adult rider, sand-colored protective suit, compact backpack, opaque bronze visor.\n\nSkimmer: one-seat rectangular solar skimmer, weathered cream shell, two dark solar panels on its nose, no wheels or logo.\n\nSky creature: colossal translucent manta-shaped cloud organism, glasslike membrane with slow internal light currents.\n\nPreserve all designs, scale relationships and travel direction left-to-right.\n\n[WORLD]\n\nPale salt canyon after a storm. Hard slope descends left-to-right toward mirrored pools. Late sun frame left; wind carries loose crystalline dust downhill.\n\n[TIMED SHOTS]\n\n0–4s — Low parallel tracking shot on Nara’s left side, camera and skimmer moving downhill at equal speed. Skimmer banks 20 degrees into one right curve; Nara leans with it. The rear edge throws one continuous dust ribbon that follows the curve and settles under gravity.\n\n4–8s — Wide aerial trailing shot, never crossing the travel axis. The glass manta glides overhead in the same direction but slower, casting a broad moving shadow that overtakes Nara. Its wings complete one heavy downstroke; canyon dust reacts half a second later.\n\n8–12s — Front three-quarter low shot moving backward at matched speed. A narrow salt ridge breaks ahead. Nara pulls the skimmer nose upward and clears one two-meter gap; suspension compresses visually on landing through a brief body dip and dense dust burst. The manta passes above, filling the sky as camera holds horizon level.\n\n[AUDIO]\n\nSteady electric turbine whine, granular salt hiss and strong crosswind. Wing downstroke produces a deep airy pressure pulse at 6s. Landing has one low impact plus settling crystals. No dialogue or music.\n\n[CONSTRAINTS]\n\nSkimmer remains in ground effect, never becomes an aircraft. Rider stays seated with both hands on controls. Physically delayed dust and shadow reaction. One gap, one jump, one creature. No collision, explosion, weapon, duplicate vehicle, changing visor, impossible camera teleport, backwards dust, logos, text or watermark.", "tags": [ "kling-4.0", "fantasy-scifi", "dialogue", "aerial", "rain", "explosion" ], "params": { "duration": "12s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/action-and-vfx.md", "tier": "reputable-3rd-party" }, "notes": "科幻追逐:每拍只做一个特技、尘埃与阴影延迟半秒反应,三镜头保持轴线,动作可读性的写作方法。" }, { "id": "k4-012", "model": "kling-4.0", "category": "portrait", "title": "Kling 4.0 未发送的语音备忘", "prompt": "[OUTPUT]\n\n12 seconds, 9:16 vertical, one uninterrupted handheld close-up, natural smartphone night footage, no cuts.\n\n[CONTINUITY ANCHOR]\n\nNoor: adult woman in her late 20s, warm brown skin, short coiled hair, oversized gray sweatshirt, tiny silver stud earrings. She sits on a kitchen floor against pale cabinets. Plain phone in her left hand; right thumb hovers above an unmarked send area.\n\n[CAMERA]\n\nPhone-camera viewpoint from a friend sitting on the opposite floor, chest-height close-up. Small breathing drift only; lens and distance remain constant. Focus stays on Noor’s eyes, then briefly shifts to her hovering thumb at 8s and returns to her face.\n\n[PERFORMANCE TIMELINE]\n\n0–3s — Noor listens to the end of her own recorded message through one earbud. Her jaw is held tight; breathing shallow.\n\n3–7s — She mouths the first word of a reply but makes no sound. Her eyes become wet without tears falling. She gives a tiny self-conscious laugh and looks away.\n\n7–10s — Her right thumb lowers toward the phone, pauses one centimeter above it, then deliberately moves away. The phone stays supported in her left palm.\n\n10–12s — Shoulders release on a slow exhale. She locks the phone with one soft click and looks directly into the camera, calm rather than defeated.\n\n[AUDIO]\n\nRefrigerator hum, faint traffic through a closed window, sweatshirt fabric and breath. A barely audible voice-note tail from the earbud is unintelligible. One phone lock click at 10.7s. No dialogue, music or notification sound.\n\n[CONSTRAINTS]\n\nNo visible app UI, readable text, tears falling, beauty filter or theatrical sobbing. Preserve face, earrings, phone, hands and floor position. Natural fingers and phone contact; no screen glow changing color; no cuts, zoom, orbit or background change.", "tags": [ "kling-4.0", "portrait", "dialogue", "close-up", "night", "rain" ], "params": { "aspect": "9:16", "duration": "12s" }, "source": { "name": "flaqai/awesome-kling-4-0", "url": "https://github.com/flaqai/awesome-kling-4-0/blob/main/prompts/cinematic-and-dialogue.md", "tier": "reputable-3rd-party" }, "notes": "一镜到底竖屏微表演:情绪按秒渐进的时间轴加焦点调度,手机夜拍质感的肖像叙事上限。" }, { "id": "mx-001", "model": "minimax-h3", "category": "dialogue-driven", "title": "韩式黑色电影预告:原生韩语对白", "prompt": "16:9, 15s, hyper-realistic Korean noir crime teaser.\n\nA rain-soaked Korean woman (late 20s, trench coat) enters an abandoned underground nightclub searching for her missing sister. A scarred crime boss (40s), presumed dead for years, emerges from the shadows with a lit cigarette. Atmosphere: 90s Korean noir, practical neon, rain, smoke, film grain, tactile realism, no CGI gloss.\n\nShots:\n\n1. She enters: \"언니... 여기 있어?\"\n\n2. Lighter ignites, revealing his scar. \"오랜만이네.\"\n\n3. She freezes: \"당신... 죽은 줄 알았는데.\"\n\n4. Slow push-in. He smiles: \"네 언니가 날 죽였다고 생각했겠지.\"\n\n5. They lock eyes. Thunder. Cut to black.\n\nAudio: Rain, thunder, jazz crackle, lighter click, intimate silence during dialogue. #MiniMaxH3", "tags": [ "minimax-h3", "dialogue-driven", "dialogue", "night", "rain", "neon" ], "params": { "aspect": "16:9" }, "source": { "name": "imaginevid/awesome-minimax-h3-prompts-and-skills", "url": "https://github.com/imaginevid/awesome-minimax-h3-prompts-and-skills", "tier": "reputable-3rd-party" }, "notes": "H3 首批原生对话验证案例:直接在 prompt 里写韩语对白并配音轨描述,展示 H3 生成精准台词+环境音的原生音频能力" }, { "id": "mx-002", "model": "minimax-h3", "category": "music-video", "title": "面无表情热舞:音频波形驱动节奏", "prompt": "Use the supplied 15-second audio slice as the exact rhythm reference. When the parcel arrives, the character's body begins dancing in perfect synchronization with the waveform: fully committed on every beat, while the face remains completely deadpan and the eyes stay locked on camera. Preserve this trigger, contrast, timing rule, and character behavior in every scene so the repeated mechanism gives the film one coherent energy.", "tags": [ "minimax-h3", "music-video", "reference" ], "params": {}, "source": { "name": "imaginevid/awesome-minimax-h3-prompts-and-skills", "url": "https://github.com/imaginevid/awesome-minimax-h3-prompts-and-skills", "tier": "reputable-3rd-party" }, "notes": "H3 独有音频 reference 用法:把 15 秒音频切片当作精确节奏参考,触发式舞蹈机制可跨场景复用\n演示视频:https://video.twimg.com/amplify_video/2083031810494324736/vid/avc1/1920x1440/ep8hqWhyscuF42MY.mp4?tag=29" }, { "id": "mx-003", "model": "minimax-h3", "category": "action", "title": "五参考追车大战:沙丘越野车逃脱无人机", "prompt": "Create a 15-second, 16:9 photoreal cinematic action sequence with native stereo audio. Treat the five images as coordinated multimodal references for identity, vehicle design, environment, performance, action, cinematography, and sound.\n\nUse Image 1 as the strict facial-identity reference for the female racer. Preserve her exact face, dark tied-back hair, amber goggles, dusty skin, black tactical suit, shoulder armor, gloves, and focused expression.\n\nUse Image 2 as the strict reference for her full costume, body proportions, armored black dune buggy, tire scale, exposed suspension, cyan headlights, roll cage, desert lighting, and industrial-outpost background.\n\nUse Image 3 as the action reference for the pursuing drones, vehicle speed, dust trails, fire, debris, chase intensity, and camera proximity.\n\nUse Image 4 as the location reference for Outpost 07, including its rusted towers, pipelines, bridges, rocky terrain, distant mountains, warm sunset atmosphere, and industrial scale.\n\nUse Image 5 as the cockpit and performance reference. Preserve the same steering wheel, open roll cage, goggles, gloves, facial identity, driving posture, and vehicle interior.\n\nSequence\n\n[0–2.5 seconds]\n\nOpen on Image 4 with an extreme-wide aerial establishing shot of Outpost 07 at sunset. The industrial complex fills the right side of frame while an empty desert route curves through the rocky foreground.\n\nThe camera dives rapidly toward the road as a small black dune buggy bursts from beneath an elevated pipeline, throwing a long dust plume behind it.\n\nKeep the vehicle moving consistently from left to right toward the outpost. No direction reversal.\n\nAudio begins with dry desert wind, distant industrial machinery, a low cinematic pulse, and the buggy engine approaching rapidly.\n\n[2.5–5 seconds]\n\nCut at peak engine sound to Image 5.\n\nTight frontal cockpit shot mounted just ahead of the driver. The vehicle shakes naturally over rough ground. Her hands hold the steering wheel firmly while she makes small, physically accurate corrections.\n\nHer amber goggles remain on top of her head. Loose strands of hair and fabric straps react to wind and vibration. Her eyes briefly check the left mirror, then return immediately to the road.\n\nA red warning light reflects across her face as a targeting alarm begins.\n\nDo not make her turn the steering wheel excessively. Her body, wheel movement, and vehicle direction must remain mechanically connected.\n\n[5–8 seconds]\n\nCut to a low front three-quarter tracking shot based on Image 3.\n\nThree pursuit drones descend behind the buggy in a triangular formation. Their rotors, stabilizers, and body movement respond realistically to speed and turbulence.\n\nThe lead drone fires into the sand beside the buggy. The impact creates a narrow eruption of dirt, sparks, and fragmented rock rather than an oversized fireball.\n\nThe racer steers sharply around the impact. The buggy's front wheels turn first, the suspension compresses, the body leans, and the rear tires slide outward before regaining traction.\n\nThe camera tracks beside the vehicle without spinning or overtaking it.\n\n[8–11 seconds]\n\nContinue the same drift into a rear three-quarter shot.\n\nThe buggy races toward a narrowing passage between a rock wall and the outer structures of Outpost 07. The racer pulls a mechanical handbrake lever for one brief moment, rotating the vehicle through the opening.\n\nOne drone follows too closely and clips a rusted overhead pipe. Its wing breaks, sending the drone tumbling into the sand behind her.\n\nShow the collision in the background while keeping the buggy dominant and moving forward. No slow motion.\n\nAudio: tire scrape, suspension impact, metal tearing, drone rotors failing, engine rev rising.\n\n[11–13 seconds]\n\nCut to a wheel-level macro shot.\n\nThe right rear tire bites into loose sand. Stones fire backward while the suspension rebounds. The camera rises naturally along the buggy's side and reveals the two remaining drones closing in.\n\nThe racer presses a guarded switch beside the steering wheel.\n\nA compact rear-mounted electromagnetic pulse discharges as a restrained blue-white distortion wave, briefly disrupting the drones' lights and stabilizers.\n\nNo magical energy, lightning storm, or giant explosion.\n\n[13–15 seconds]\n\nCut to a low frontal hero shot as the buggy clears the outpost gate at full speed.\n\nThe two disabled drones fall into the dust behind it while the vehicle launches from a shallow ridge. Keep the jump low, heavy, and physically believable.\n\nDuring the brief airborne moment, cut to Image 1 for a tight close-up of the racer. Preserve her exact identity as warm firelight and cool dashboard light cross her face. Her expression remains controlled and determined.\n\nThe buggy lands hard beyond the gate. The suspension compresses, the engine roars, and the vehicle continues directly into the desert.\n\nEnd with a sharp cut to black on the landing impact.\n\nVisual Direction\n\nPremium live-action science-fiction action trailer with realistic CGI integration, warm orange sunset light, restrained steel-blue technology accents, dusty atmosphere, hard surface detail, cinematic contrast, subtle film grain, realistic motion blur, and 24 FPS movement.\n\nUse wide shots to establish scale, cockpit close-ups for tension, low tracking shots for speed, and mechanical macro shots for physical detail.\n\nKeep the editing fast but readable. Every shot must begin from the physical state established by the previous shot.\n\nAudio Direction\n\nNative stereo sound with:\n\nAggressive combustion engine\n\nTire friction over sand and rock\n\nSuspension rattles and chassis vibration\n\nDrone rotors and targeting alarms\n\nSand impacts, metal collisions, and falling debris\n\nLow percussion and rising electronic tension\n\nOne heavy bass impact on the final landing\n\nKeep music underneath the vehicle and environmental sounds. No dialogue or voice-over.\n\nRestrictions\n\nNo subtitles, titles, logos, watermarks, additional racers, pedestrians, creatures, motorcycles, futuristic city skyline, nighttime transition, costume changes, facial changes, vehicle redesign, additional wheels, floating vehicle parts, distorted hands, incorrect steering, reversed wheel rotation, teleportation, random explosions, oversized fireballs, weightless motion, impossible jumps, camera spins, circular camera moves, rapid zooms, fluid morphs, soft dissolves, or changes in travel direction. #MiniMaxH3", "tags": [ "minimax-h3", "action", "dialogue", "reference", "close-up", "aerial" ], "params": { "aspect": "16:9" }, "source": { "name": "imaginevid/awesome-minimax-h3-prompts-and-skills", "url": "https://github.com/imaginevid/awesome-minimax-h3-prompts-and-skills", "tier": "reputable-3rd-party" }, "notes": "H3 多模态 reference 范式标杆:5 张图片各司其职(身份/载具/动作/场景/座舱)+ 原生立体声音轨描述,附结果视频" }, { "id": "mx-004", "model": "minimax-h3", "category": "image-to-video", "title": "孟买季风 FPV:手绘路线图引导航拍", "prompt": "Reference: Use the attached @ Image1 1 as the exact first frame and environment reference. The green drawn line is only the camera flight-path guide. Do not show the green line in the final video.\n\nCreate a 15-second hyper-real cinematic FPV drone video over Mumbai in heavy monsoon weather, using the same skyline, interchange, sea edge, road geometry, dense buildings, rainy morning atmosphere, grey clouds, and wet surfaces from the reference image. Keep everything realistic and grounded. The drone must follow the green guideline path as a smooth continuous move: begin near the left side of the interchange, rise and sweep across the middle highway corridor, continue through the right-side city edge, then climb higher into the skyline, perform a tight spiral/orbit near the upper tower cluster, descend back toward the interchange, curve around the large circular loop, and finally glide out toward the lower-right water edge. The move should feel fast, elegant, immersive, and controlled — like a premium FPV drone shot, not a random flyover.\n\nTIMELINE\n\n 0:00–0:03 Start low near the left feeder ramp above the wet interchange. Heavy rain falls across the lens. The drone accelerates forward and slightly upward, revealing slick roads, tiny moving traffic, dark sea barriers, and the circular loop ahead.\n\n0:03–0:06 Follow the path upward across the mid-frame highway corridor. Keep the skyline growing larger. Wet roads shimmer, rain haze softens the distance, and the sea remains visible to the right. Motion is quick but readable.\n\n0:06–0:09 Sweep along the right-side city edge and climb higher. Dense towers, mid-rise blocks, and dark green patches below feel soaked and monsoon-heavy. The camera begins rising toward the upper skyline.\n\n0:09–0:11 Reach the upper skyline cluster and perform a tight spiral/orbit around the central high-rises. Storm clouds loom above, rain streaks rush past, and the city beneath feels deep, layered, and cinematic.\n\n0:11–0:13 Exit the spiral and descend in a smooth arc back toward the interchange. The circular loop becomes dominant again. Maintain realistic drone inertia and readable geography.\n\n0:13–0:15 Curve around the large circular flyover loop and finish with a fast glide toward the lower-right over the water, ending on a dramatic monsoon reveal of wet infrastructure, dark sea, and rain-washed Mumbai.\n\nCAMERA LANGUAGE FPV drone POV, smooth stabilized flight, realistic banking turns, slight motion blur from speed and rain, believable acceleration and inertia, one continuous shot, no teleporting, no abrupt cuts, no impossible passes through buildings.\n\nVISUAL STYLE Hyper-real cinematic realism, heavy monsoon rain, grey storm clouds, wet roads, reflective flyovers, rain haze, cool blue-grey morning light, dark sea, atmospheric depth, realistic Mumbai scale, premium documentary-film look.\n\nAUDIO No voiceover. No dialogue. Use heavy rainfall, rushing wind, distant thunder, soft drone motor whirr, faint traffic hum, sea movement, and subtle cinematic tension music building with the motion.\n\nIMPORTANT Do not show the green route line. No text, captions, logos, or watermark. Keep Mumbai realistic, monsoon-heavy, and geographically believable.", "tags": [ "minimax-h3", "image-to-video", "dialogue", "reference", "drone", "rain" ], "params": {}, "source": { "name": "imaginevid/awesome-minimax-h3-prompts-and-skills", "url": "https://github.com/imaginevid/awesome-minimax-h3-prompts-and-skills", "tier": "reputable-3rd-party" }, "notes": "巧妙的图生视频技巧:在参考图上画绿色引导线定义无人机航线,明确要求隐藏路线线——路线可控性范式" }, { "id": "mx-005", "model": "minimax-h3", "category": "comedy", "title": "1980s 复古机器人家庭喜剧:开场图转视频", "prompt": "Use the supplied image as the exact opening frame. Create a hilarious, high-budget 1980s live-action family comedy movie scene, photographed on a real soundstage with practical robot costumes, animatronics, handmade props, vintage wardrobe, and authentic 35mm film color.\n\nThe entire group suddenly hears something above them. Everyone's eyes dart upward at the same moment. The men, women, children, and robots dramatically tilt their heads back, point toward the ceiling, and erupt into chaotic celebration. They gasp, scream, laugh, jump, wave their arms, slap each other on the shoulder, and completely lose their minds with exaggerated 1980s comedy reactions. The children bounce excitedly. The adults stumble into one another. The robots flash their eyes, flap their mechanical arms, spin clumsily, and celebrate with funny practical animatronic movements.\n\nOne man shouts:\n\n“OPEN SOURCE?!”\n\nThe entire crowd answers together:\n\n“MINIMAX H3!”\n\nThey cheer wildly as one robot attempts a victory dance, loses its balance, and is caught by the shocked family at the final second.\n\nFast, energetic comic timing with believable ensemble choreography. Begin with a brief locked group shot, then a subtle handheld push-in as the chaos escalates. Keep every original person, robot, shirt design, and the large headline clearly recognizable. Preserve facial identity, wardrobe, composition, lighting, and 1980s production design.\n\nAuthentic optical softness, warm tungsten lighting, rich film grain, slight gate weave, practical effects, natural motion blur, expressive physical acting, polished studio-comedy sound mix, triumphant synthesizer sting, cheering, robot beeps, and comedic percussion.\n\nNo CGI, no digital-looking robots, no morphing, no extra people, no duplicated characters, no distorted faces, no rewritten text, no spelling changes, no modern clothing, no style shifts, and do not turn the scene into animation.", "tags": [ "minimax-h3", "comedy", "rain", "dance", "product", "crowd" ], "params": {}, "source": { "name": "imaginevid/awesome-minimax-h3-prompts-and-skills", "url": "https://github.com/imaginevid/awesome-minimax-h3-prompts-and-skills", "tier": "reputable-3rd-party" }, "notes": "图生视频喜剧范例:锁定开场图身份/服装/文字,靠群戏调度+喜剧音效设计驱动笑点,展示 H3 原生音频与群戏一致性" }, { "id": "mx-006", "model": "minimax-h3", "category": "music-video", "title": "七色 K-pop 回归预告:精准排版字效", "prompt": "Create a 15-second, 16:9, 2K, high-budget K-pop girl-group comeback teaser. The group consists of seven original Korean female idols, all aged 22 or older; they must not resemble real celebrities. Give each member a distinct face, hairstyle, outfit, and signature color while maintaining a photorealistic live-action appearance.\n\nMembers: MIA uses purple-and-black styling with violet electricity. IVY uses emerald-and-black styling with green luminous vines. NIA uses ice-blue styling with glass fragments and icy mist. IRIS uses silver-and-white styling with iridescent light. MAY uses black-and-red styling with red sparks. ARI uses gold-and-black styling with golden particles. XENA uses deep-blue styling with blue lasers. Their initials spell MINIMAX.\n\nOnly the following text may appear: MIA, IVY, NIA, IRIS, MAY, ARI, XENA, and MINIMAX. Present each name as premium character-introduction typography with condensed uppercase letters, geometric cuts, metallic outlines, translucent echoes, scan lines, small HUD circles, letter fragments, and energy trails. Keep every name accurate and legible without covering a face.\n\nTimeline: 0.0-0.5s, seven colored light trails sweep across black and form an incomplete MINIMAX outline. 0.5-2.0s, MIA in a violet mirrored corridor as the camera pushes in. 2.0-3.5s, violet electricity transforms into green vines and IVY emerges. 3.5-5.0s, green light freezes into ice-blue glass and NIA emerges through cold mist. 5.0-6.5s, glass refracts into iridescent light and IRIS turns in a silver-white mirrored space. 6.5-8.0s, red sparks ignite and MAY swings translucent red fabric toward camera. 8.0-9.5s, sparks become gold particles and ARI walks out from a black-and-gold stage. 9.5-11.0s, the gold ring becomes a deep-blue laser corridor and XENA steps forward. Briefly decelerate each introduction for 0.3s while hair, clothing, mist, and light continue moving; hold each name for at least 0.5s.\n\nFrom 11.0-15.0s, a blue laser reveals a large black mirrored stage. All seven members enter from different directions and form a clear three-front, four-back formation with every face visible. Pull back slowly from a low angle. Contract each name to its first letter, M I N I M A X, then combine the letters into a large MINIMAX title built from structural lines, geometric cuts, metallic fill, translucent echoes, and a final seven-color sweep. Hold the group looking into camera for one second. Use native stereo sound: electronic bass, percussion, electricity, glass, airflow, sparks, lasers, and typographic lock-on effects. No dialogue. Avoid celebrities, minors, gibberish, misspellings, extra text, text over faces, identity changes, similar-looking members, revealing outfits, anime, game CGI, cheap neon, excessive particles, and unnecessary 3D rotation.", "tags": [ "minimax-h3", "music-video", "dialogue", "neon", "anime" ], "params": { "aspect": "16:9" }, "source": { "name": "imaginevid/awesome-minimax-h3-prompts-and-skills", "url": "https://github.com/imaginevid/awesome-minimax-h3-prompts-and-skills", "tier": "reputable-3rd-party" }, "notes": "H3 文字渲染能力的硬核演示:7 位成员名+MINIMAX 标题逐字排版,每段 0.3s 减速持名 0.5s,七色光效随音乐变奏" }, { "id": "mx-007", "model": "minimax-h3", "category": "product-commercial", "title": "抹茶新手礼盒晨间 UGC 广告", "prompt": "Use the uploaded reference image as the exact character reference. Preserve her facial identity, hairstyle, eye color, skin tone, body proportions, and facial consistency throughout every shot.\n\nCreate a 15-second ultra-realistic UGC-style lifestyle advertisement featuring a realistic young woman sharing her favorite premium Matcha Starter Kit in a bright, minimalist Japanese-inspired kitchen with warm morning sunlight, natural wood textures, soft neutral décor, ceramic tableware, and a calm luxury aesthetic.\n\nShe smiles naturally while holding an elegant matcha tin and says this has become her favorite way to start every morning. She opens the premium matcha container, carefully scoops vibrant green matcha powder into a handcrafted ceramic bowl, pours warm water, and whisks it smoothly using a traditional bamboo whisk until a rich, creamy foam forms. She lifts the bowl, takes a small sip, closes her eyes with a genuine smile, and enjoys the peaceful morning moment.\n\nCapture authentic handheld smartphone shots, over-the-shoulder angles, cinematic macro close-ups of the vivid green matcha powder, the bamboo whisk creating silky foam, the handcrafted ceramic bowl, rising steam, natural hand movements, and warm sunlight reflecting across the table. Finish with her holding the completed bowl beside her face, smiling warmly into the camera.\n\nAuthentic creator-style content, natural facial expressions, healthy glowing skin, cozy morning atmosphere, luxury lifestyle aesthetic, subtle depth of field, premium color grading, smartphone-shot realism, vertical 9:16, 4K HDR, no text, subtitles, logos, or watermarks.", "tags": [ "minimax-h3", "product-commercial", "reference", "close-up", "one-shot" ], "params": {}, "source": { "name": "imaginevid/awesome-minimax-h3-prompts-and-skills", "url": "https://github.com/imaginevid/awesome-minimax-h3-prompts-and-skills", "tier": "reputable-3rd-party" }, "notes": "H3 的参考身份锁定做带货 UGC:角色身份严格锁定+抹茶打制触感特写,手持手机美学对电商内容有直接参考价值" }, { "id": "mx-008", "model": "minimax-h3", "category": "dialogue-driven", "title": "纸鸟避雨计划:双语音参考定音配对白", "prompt": "Create a tactile stop-motion paper-craft scene beginning from Image 1. Preserve exactly two characters: the broad indigo bird stays screen left with one folded wing held close; the smaller saffron bird stays screen right with a long pointed tail. Preserve paper colors, fold patterns, fiber texture, scale difference, seed tray, warm lantern, greenhouse structure, rain direction, and open door at far right.\n\n0–3 seconds: establishing medium-wide shot with Audio 3 prominent. A gust rattles the greenhouse; both birds look toward the right-hand door, then at each other. No dialogue. 3–7 seconds: close on the indigo bird using Audio 1, low and steady: “The roof will hold if we brace the west hinge.” Its beak movement is minimal and synchronized; the saffron bird listens without speaking. 7–11 seconds: reverse close-up using Audio 2, quick but clear: “Then I'll tie the twine. You keep the lantern dry.” The indigo bird listens and lowers its tense shoulder fold. 11–15 seconds: return to the two-shot. They each take one end of a loose natural-fiber twine and move together toward the exit; the lantern remains inside. End before they enter the rain.\n\nUse slight stop-motion cadence, consistent eye lines, and one speaker at a time. Sound hierarchy: dialogue first, rain and wooden creaks below, no music. Avoid extra birds, human hands, realistic feathers, changing folds, swapped voices, subtitles, speech bubbles, a repaired roof not yet shown, exaggerated panic, franchise resemblance, logos, or watermark.", "tags": [ "minimax-h3", "dialogue-driven", "dialogue", "close-up", "rain" ], "params": { "duration": "3s" }, "source": { "name": "aivideoweb/awesome-minimax-h3-prompts", "url": "https://github.com/aivideoweb/awesome-minimax-h3-prompts/blob/HEAD/prompts/21-character-dialogue-performance.md", "tier": "reputable-3rd-party" }, "notes": "H3 音视频统一 reference 的教科书:图像定角色场景,Audio 1/2 分别授权两个角色的音色,Audio 3 只管雨声环境,一人一句配对口型" }, { "id": "mx-009", "model": "minimax-h3", "category": "technique-snippet", "title": "相机语法迁移:借运镜节奏不借画面", "prompt": "Create an original studio advertisement for the fictional graphite-and-cork portable speaker in Image 1. Transfer only the abstract camera grammar from Video 1: shot lengths, direction of travel, acceleration curve, focal transitions, and cut timing. Do not copy its subject, setting, colors, props, lighting design, composition, campaign idea, text, or recognizable effects.\n\n0–3 seconds: slow three-quarter push toward the full speaker on a charcoal plinth; amber rim light reveals the cork edge. 3–6 seconds: cut on Audio 1's second downbeat to a controlled lateral macro move across the cork pores and metal seam. 6–9 seconds: repeat Video 1's orbit speed around the circular woven grille while the product remains stationary and physically intact. The amber status light turns on once at the final beat of this section. 9–12 seconds: pull back to the original three-quarter product view; a shallow pattern of light crosses the acoustic panel behind it and settles, leaving clean negative space for typography added later.\n\nPreserve the exact silhouette, circular grille, cork panels, seam placement, amber indicator, scale, and material response from Image 1. Sound intent: Audio 1 plus restrained tactile clicks and low room tone; do not imply measured loudness or acoustic performance. Avoid logos, generated copy, extra controls, duplicate speakers, pulsing geometry, visible sound waves, floating product, copied reference-video scenery, or watermark.", "tags": [ "minimax-h3", "technique-snippet", "reference", "rain", "product" ], "params": { "duration": "3s" }, "source": { "name": "aivideoweb/awesome-minimax-h3-prompts", "url": "https://github.com/aivideoweb/awesome-minimax-h3-prompts/blob/HEAD/prompts/20-multireference-camera-transfer.md", "tier": "reputable-3rd-party" }, "notes": "H3 视频 reference 的新玩法:只迁移参考视频的抽象运镜语法(景别/加速度/切点),不复制其内容——解决版权风险的实用技巧" }, { "id": "mx-010", "model": "minimax-h3", "category": "music-video", "title": "车站双语二重唱:中西歌声对答", "prompt": "Create an intimate, fictional musical exchange on an almost-empty covered station platform at blue hour. Preserve Performer A from Image 1 in a charcoal coat on screen left and Performer B from Image 2 in a rust scarf on screen right. Preserve the platform roof and bench geometry from Image 3; no real rail operator marks.\n\n0–4 seconds: medium two-shot, distant train hum. Performer A sings one soft Mandarin line, “下一站,我们还会再见,” with a calm, hopeful expression. Performer B listens without lip movement. 4–8 seconds: cut on a station-light flicker to Performer B, who answers in Spanish, “La próxima vez, llegaré temprano.” Performer A now listens. 8–12 seconds: return to the two-shot as they sing a wordless two-note harmony; a train light passes behind them and the camera gently pushes in.\n\nKeep exact voice-to-face ownership, natural breathing, intelligible pronunciation, stable eyelines, consistent platform geography, and modest performance. Sound intent: voices foreground, light rail ambience, no additional lyrics, no non-diegetic music. Avoid translated subtitles baked into the image, overlapping mouth movement, voice swapping, melodramatic gestures, real transit logos, or watermark.", "tags": [ "minimax-h3", "music-video", "rain" ], "params": { "duration": "4s" }, "source": { "name": "aivideoweb/awesome-minimax-h3-prompts", "url": "https://github.com/aivideoweb/awesome-minimax-h3-prompts/blob/HEAD/prompts/13-music-performance-audio.md", "tier": "reputable-3rd-party" }, "notes": "H3 多语音色绑定的音乐用法:中文+西班牙语两段歌声各有授权音色,规定发声者归属与聆听者闭嘴——双语演唱的可复现模板" }, { "id": "se-001", "model": "seedance-2.5", "category": "dialogue-driven", "title": "出租车分手戏:微表情连续情绪演变", "prompt": "人设女性:年轻漂亮,长发披肩,坐在出租车后排。角色状态:与男友分手告别,表面强撑微笑、内心悲伤压抑,最终情绪崩溃嚎啕大哭。场景出租车内部的后排车厢作为空间环境参考,前排副驾驶空位(前景局部失焦虚化);侧车窗与后车窗均被雨水打湿;车窗外为城市街道夜景、霓虹灯光、淅淅沥沥的小雨。画面风格真人实拍质感的青春恋爱电视剧,现代都市写实服化道,细腻情感戏审美,情绪戏节奏克制。影调车内整体偏暗,窗外城市夜景与霓虹灯光透过雨打车窗,在车内形成流动的彩色光斑与明暗切片,勾勒女性面部轮廓与长发边缘;车窗玻璃上雨珠反光;夜景冷调为主,霓虹光斑冷暖交织。全局约束无字幕、禁止生成字幕。画面内容分镜\n\n1,0s-6s,车内中景,固定机位,摄像机位于前排副驾驶位置朝后方拍摄;前景左下方为副驾驶空座椅背局部(失焦虚化),女性位于画面右侧三分线位置。她侧着脸望向右侧车窗外,轻咬下唇,微笑着与窗外人道别,窗外站着一名身穿灰色夹克黑色牛仔裤的男性,透过车窗只能看见男性的局部身体:对白:女性(轻咬下唇,微笑):「就这样,不用送了。」随后她转头面向镜头,对前排司机说:对白:女性:「司机,开车。」说完她不经意地带出一声压抑的细微啜泣,随即轻轻舒了一口气,脸上仍保持着刚才的微笑。出租车随即启动,车身轻微震动一下,车辆起步向前开动,窗外男性朝着车窗挥手再见(局部入画),男性与窗外雨幕与霓虹灯光快速向后掠过。音效:淅沥雨声打在车顶与车窗,城市夜晚低缓的环境底噪;对白期间人声清晰突出;一声压抑的细微啜泣;轻舒气声;发动机启动声与车身轻微震动;轮胎碾过湿路面的起步声。分镜\n\n2,6s-15s,切换车内后排近景,手持拍摄,轻微仰拍,镜头随行驶中的车身轻微晃动;焦点在女性肩部以上正面,她背后隔着被雨打湿而模糊的后车窗,城市街景与霓虹灯光逐渐远去。一开始女性脸上仍带着微笑,没有皱眉也没有哭,慢慢地她开始轻微皱起眉头,小幅摇了摇头,像是反悔刚才说出口的话;接着她的嘴角微微下压,抿紧双唇,眼眶开始泛红;她稍稍仰起头试图忍住,眼神有些游离地瞥向旁边车窗外;接着她终于再也压抑不住——嘴角明显下垂,眉毛内收收紧,鼻子开始抽搐,眼神黯淡,眼眶湿润,泪水开始流出;然后她垂下头、微微蜷缩着肩膀,随即猛然仰头,嚎啕大哭,一边哭一边抬手擦去脸上止不住的泪水。整段为逞强微笑→压抑情绪→悲伤外泄→情绪大爆发的连续情绪变化,一镜到底。音效:雨声持续;行驶中发动机低鸣与轮胎碾过湿路面的沙沙声;呼吸渐重、吸鼻声;压抑啜泣转为放声大哭;手背擦拭泪水的摩擦声与衣料声。全局配乐雨夜分手情绪戏(15s 配乐轨,无歌词)0-6s 无BGM,只保留对白和环境声;6-8s 微笑→轻微皱眉 | 钢琴单音/两音动机从雨声里渗出,极弱 pad 垫底,像回忆浮上来;8-10s 摇头反悔→抿唇→眼眶泛红 | 钢琴变叹息式下行音型,大提琴长音进入;10-12s 仰头忍泪→眼神游离 | 小提琴组加入,和声加厚,钢琴分解和弦铺开;12-13.5s 嘴角下压→鼻翼抽动→泪落 | 情绪临界:弦乐揉弦+力度加强,钢琴和弦加重;13.5-15s 垂头蜷缩→仰头嚎啕大哭 | 段落峰值:弦乐齐奏主题,钢琴八度重音。", "tags": [ "seedance-2.5", "dialogue-driven" ], "params": {}, "source": { "name": "kafanxx/awesome-seedance", "url": "https://github.com/kafanxx/awesome-seedance/blob/HEAD/docs/gallery-seedance-2-5-part-4.zh.md", "tier": "reputable-3rd-party" }, "notes": "Seedance 2.5 微表情断档领先的实测案例:15 秒一镜到底完成逞强微笑→崩溃大哭的连续情绪变化,还附了配乐时间轴\n演示视频:https://goodcase.ai/cases/seedance-2-5-8d136b59e95a" }, { "id": "se-002", "model": "seedance-2.5", "category": "fantasy-scifi", "title": "一张照片飞天:科幻史诗升空短片", "prompt": "中国人能飞✈️\n\n想飞吗?只需输入自己的照片,加上下面的提示词,你也能飞✈️😄\n\nSeedance 2.5 prompt 👇\n\n科幻史诗真人电影(Sci-fi Epic Live-Action),ARRI Alexa 35 机身 + Panavision Ultra Vintage 变形宽银幕镜头(Anamorphic),2.39:1 宽银幕画幅,Tiffen Black Pro-Mist 1/4 柔光镜,Kodak 250D 胶片质感、可见但细腻的颗粒,高光边缘带红橙渗色光晕(Halation),180°快门角自然运动模糊,暗部保留细节不死黑,8K超清,真实摄影(Photorealistic),无动漫感、无CG塑料感\n\n30秒\n\n从地球荒原一路上升,穿雷暴云 → 平流层 → 近地轨道 → 月球 → 小行星带 → 太阳表面\n\n主角(全片同一人,服装与外形完全以参考图为准)\n\n[00:00-00:02] 镜头1:负手而立(85mm 长焦 / Slow Dolly In)\n\n广角远景:主角@图片1背对镜头站在枯黄草原上,双手背在身后,一动不动。天顶被一整片巨大的乳状积云压满,远处地平线只有一条明亮的逆光带。\n\n画面:风吹得草浪成片倒伏,逆光把草叶边缘描出一圈毛茸茸的金边。\n\n机位:85mm 长焦压缩空间,镜头缓慢向前推进。\n\n[00:02-00:04] 镜头2:起飞冲击(40mm / Speed Ramp + Whip Tilt Up)\n\n0.3秒慢动作:主角@图片1脚下一圈环形冲击波贴地扩散出去,方圆几十米的草被压平成一个同心圆,碎草和沙尘垂直腾起。\n\n随即切全速:他猛然垂直升空,画面只剩他的腿从上方出画。镜头急速上摇跟随,被气浪震了一下(Camera Shake),穿过尘土冲进云层。\n\n[00:04-00:07] 镜头3:雷暴云中心(40mm)\n\n主角@图片1缩成一个小黑点在云柱中垂直上升,负手姿态没有改变。四周云里的蓝白色闪电一道道炸开,每次闪把周围的云照亮一瞬,他的剪影在亮起的云幕上被压成一个黑点。\n\n机位:跟随上升,云层向画面下方飞速倒退,变形镜头把画面边缘轻微拉长。\n\n[00:07-00:09] 镜头4:冲出云顶(50mm / Backlit Emergence)\n\n主角@图片1逆光穿出金色云层,身后的云被撞出一个漏斗形的空洞,暖金色光晕从他背后爆开。\n\n画面:强逆光在他轮廓外缘烧出一圈红橙色渗色光晕(Halation),高光被柔光镜化开但不过曝。\n\n[00:09-00:13] 镜头5:平流层360°环绕(50mm / Full 360° Orbit)\n\n画面:主角@图片1在云海上方静止悬停,双手垂在身侧一动不动,全程不转身。天空上半部已是深棕黑的临近太空,下方是被夕阳染成金色的积云海。\n\n机位(00:09-00:12):50mm 镜头从他背后起,绕着他匀速环绕一整圈——依次经过他的侧面、正面、另一侧,最后转回他的背后。环绕全程机位与他的距离保持不变,他在画面里的大小也不变。\n\n机位(00:12-00:13):镜头停在他背后不动,他从这个背面视角猛然加速垂直上升,瞬间缩成一个小点飞离画面,只剩空荡荡的金色云海。\n\n节奏:这一段要慢,是全片的呼吸点。\n\n[00:13-00:15] 镜头6:地球轨道冲刺(40mm / Side Tracking + Vapor Cone)\n\n硬切超近景:主角@图片1俯冲飞行姿态,身体水平前倾、手臂前伸,衣物被气流抽打得剧烈变形,布料褶皱清晰。背景是地球弧面和一整个台风漩涡云系。\n\n特效:他身体周围炸开一圈白色音爆云环。", "tags": [ "seedance-2.5", "fantasy-scifi", "dance" ], "params": {}, "source": { "name": "goodcase.ai", "url": "https://goodcase.ai/cases/johnagi168-seedance-ai-8424f361fbe0", "tier": "unverified-3rd-party" }, "notes": "中文社区火爆的 2.5 模板:四段式中文格式,单张照片换脸飞行,30 秒跨度荒原→太阳\n演示视频:https://goodcase.ai/cases/johnagi168-seedance-ai-d704aa5a63ac" }, { "id": "se-003", "model": "seedance-2.5", "category": "sports", "title": "偶像女孩 HIIT 训练日志:DV 磁带机实录", "prompt": "CAMERA: DV 16mm tape camcorder, handheld POV/self-shot by CHASE. Occasional propped shots on bench/floor. Shaky framing, soft focus, clumsy zooms, face cutoffs, chaotic movement during exercise, steadier during recovery. Camcorder never visible.\n\nLOOK: Soft blurry tape quality, faint noise, bloomed gym lights, flickering exposure, muted contrast, realistic skin tones. Sweat progressively builds.\n\nSTYLE: High-energy, breathless HIIT vlog. Intense exercise bursts → short recovery beats. Genuine exhaustion and determination.\n\nCHASE: Korean idol in her 20s, long black high ponytail becoming damp, expressive eyes, flushed cheeks, slim athletic build. Modest long-sleeve athletic top, joggers/leggings, sneakers, towel, no jewelry.\n\nSETTING: Evening gym open floor, mats, mirrors, small speaker/phone, water bottle, soft overhead lighting.\n\n30s / 9 CUTS:\n\n1. 3s Propped: Ready on mat, bouncing. “Okay, HIIT round thirty seconds on, let’s go!”\n2. 3s Handheld: Fast jumping jacks. “Okayalready feeling this”\n3. 2.5s Propped: Hands on knees, recovering. “Okay five seconds that’s it”\n4. 3s Low/propped: Intense mountain climbers, sweat visible. Heavy breathing only.\n5. 2.5s Propped: Sitting back, wiping sweat. “Okay, that one’s brutal”\n6. 3.5s Handheld: Burpees, increasing motion blur. “Why did Ipick burpees”\n7. 2.5s Macro: Hands gripping mat, sweat dripping, chest heaving. No dialogue.\n8. 3s Propped: Collapses onto back, laughing breathlessly. “Okay I think I’m dying”", "tags": [ "seedance-2.5", "sports", "dialogue" ], "params": {}, "source": { "name": "kafanxx/awesome-seedance", "url": "https://github.com/kafanxx/awesome-seedance/blob/HEAD/docs/gallery-seedance-2-5-part-3.zh.md", "tier": "reputable-3rd-party" }, "notes": "2.5 长单镜头+多分镜能力:30 秒 9 个切镜的运动 vlog,DV 磁带机质感与汗水渐进细节展现物理一致性\n演示视频:https://goodcase.ai/cases/doctorwasif-seedance-ai-653b4911153b" }, { "id": "se-004", "model": "seedance-2.5", "category": "quiet-moments", "title": "深夜厨房手机实录:做饭 30 秒", "prompt": "FORMAT: 30-second horizontal 16:9 realistic smartphone video.\nCreate an authentic, imperfect everyday late-night kitchen phone recording, as if a friend casually left a phone recording on the counter. Use a standard smartphone camera with mild autofocus hunting, subtle exposure shifts, natural motion blur, compression artifacts, and slightly uneven handheld framing. No polished commercial look.\n00–05s — WALKING IN\nThe character walks into a quiet home kitchen carrying a small grocery bag. They set it on the counter, sigh quietly, and pull out a few simple ingredients. They glance at the phone for a second before turning back toward the counter.\nCamera remains several feet away at counter height, as if casually recording from a nearby surface.\n05–10s — GETTING STARTED\nThey wash their hands, turn on the kitchen light, and begin preparing the ingredients.\nA cutting board shifts slightly against the counter. They pause to move it into a better position.\nThe character looks tired but relaxed, occasionally brushing their hair away from their forehead.\n10–15s — COOKING\nThey place a pan on the stove and begin cooking.\nSteam gradually becomes visible. The phone autofocus briefly shifts between the character and the foreground pan before settling again.\nThey stir the food, taste a tiny amount, think for a moment, then add a little more seasoning.\nNo readable labels or packaging should be visible.\n15–20s — SMALL MISTAKE\nThe character gets distracted for a moment and notices the food starting to cook more quickly than expected.\nThey immediately lower the heat and give a small amused shake of their head.\nThey quietly mutter something like:\n“Yeah… almost forgot about that.”\nThe delivery should feel spontaneous and understated.\n20–25s — FINISHING\nThey finish preparing the food, turn off the stove, and carefully transfer the meal onto a plate.\nThey wipe a small splash from the counter with a paper towel.\nNatural movements only—nothing overly choreographed.", "tags": [ "seedance-2.5", "quiet-moments", "night", "food" ], "params": { "aspect": "16:9" }, "source": { "name": "kafanxx/awesome-seedance", "url": "https://github.com/kafanxx/awesome-seedance/blob/HEAD/docs/gallery-seedance-2-5-part-2.zh.md", "tier": "reputable-3rd-party" }, "notes": "2.5 的反精致真实感美学:明确要求自动对焦漂移、压缩伪影、手持不稳和小失误——让 AI 视频摆脱广告味的关键 prompt 写法\n演示视频:https://goodcase.ai/cases/seedance-2-5-eba905fedcff" }, { "id": "ve-021", "model": "veo-3", "category": "dialogue-driven", "title": "Veo 3.1 原生对白与口型同步写法(120ms lip-sync)", "prompt": "Dialogue Structure: \"Character [Description] says '[Exact Quote],' [Tone Adjective], [Sync Timing].\"\n\nExample: A middle-aged detective says 'I told you not to come here,' weary and gravelly tone, 120ms lip-sync delay.\n\nUse quotation marks for spoken lines: putting the exact dialogue in quotes (says, \"...\") gives the model a clear start and stop point. List ambient sounds (like rain tapping or cafe murmur) at the end of the prompt so the engine does not blend environmental noise into the main speech track.", "tags": [ "veo-3", "dialogue-driven", "dialogue", "lip-sync", "rain" ], "params": {}, "source": { "name": "www.atlascloud.ai", "url": "https://www.atlascloud.ai/blog/ai-updates/veo-3.1-prompt-guide", "tier": "unverified-3rd-party" }, "notes": "Veo 3.1 的 48kHz native audio 对白标准写法:角色描述+says+精确引语+语气+lip-sync 时延;与已有 Veo 3 对白条目不同,这是 3.1 专属的同步语法。\n演示视频:https://www.youtube.com/embed/48jtRE42Bw0" }, { "id": "ve-022", "model": "veo-3", "category": "technique-snippet", "title": "Veo 3.1 85mm 跟拍 + 脚步重量物理与音效绑定", "prompt": "Cinematography: 85mm medium tracking shot, smooth eye-level dolly backward, shallow depth of field.\n\nSubject & Action: A handsome man... walks forward... His boots firmly contact the rain-slicked asphalt with clear weight and friction...\n\nEnvironment: Rain-soaked cinematic urban alleyway, vibrant neon lights...\n\n(原文含省略号为来源作者删节) Native audio note: explicit physical-contact wording (boots firmly contact... with clear weight and friction) forces the physics engine to render grounded weight and lets the native audio engine snap footstep sounds to the visual contact frames.", "tags": [ "veo-3", "technique-snippet", "rain", "neon" ], "params": {}, "source": { "name": "www.atlascloud.ai", "url": "https://www.atlascloud.ai/blog/ai-updates/veo-3.1-prompt-guide", "tier": "unverified-3rd-party" }, "notes": "新技巧:用物理接触动词(firmly contact with clear weight and friction)消除“太空步”并实现 action-sound binding;机位参数前置锁定空间一致性。\n演示视频:https://www.youtube.com/embed/I6RMl43E37I" }, { "id": "ve-023", "model": "veo-3", "category": "social-viral", "title": "吉祥物慢步入场(便利店夜晚病毒版)", "prompt": "A short cinematic video clip: the camera holds steady on the subject standing outside a dimly lit convenience store at night. After a beat, a tall, slender, pastel-purple cartoon rabbit-like mascot character with oversized round eyes, a wide friendly cartoon grin, and smooth, toy-like 3D-rendered proportions slowly walks into frame from just outside the light, taking three or four unhurried steps until it comes to a stop beside the subject, its shadow growing and shifting realistically as it moves under the same storefront light source. Quiet, deliberate pacing, no jump cuts, no distorted motion, no readable signage text, no logos.", "tags": [ "veo-3", "social-viral", "night" ], "params": {}, "source": { "name": "miraflow.ai", "url": "https://miraflow.ai/blog/digital-circus-cartoon-real-life-ai-photo-trend-prompts-2026", "tier": "unverified-3rd-party" }, "notes": "2026 年 9 月 TikTok/Reels 病毒趋势 Digital Circus 的视频版 prompt(通用型,可用于 Veo 3/Sora 2);写法要点:卡通角色的阴影随步伐实时变化、光影与真人严格匹配。" }, { "id": "ve-024", "model": "veo-3", "category": "social-viral", "title": "自拍转场:吉祥物突然现身", "prompt": "A short cinematic video clip: the subject holds up a phone taking a normal selfie in an ordinary daylight room, alone in frame. The camera then pushes in slightly and, in one smooth transition, a tall, slender, pastel-purple cartoon rabbit-like mascot character with oversized round eyes, a wide friendly cartoon grin, and smooth, toy-like 3D-rendered proportions is revealed standing directly behind the subject in the reflection or background where nothing was visible a moment before. Keep the lighting on the mascot consistent with the room's existing light source through the transition, smooth and deliberate pacing, no jarring cuts, no distorted motion, no readable text, no logos.", "tags": [ "veo-3", "social-viral" ], "params": {}, "source": { "name": "miraflow.ai", "url": "https://miraflow.ai/blog/digital-circus-cartoon-real-life-ai-photo-trend-prompts-2026", "tier": "unverified-3rd-party" }, "notes": "同一病毒趋势的另一视频版:一镜式自拍 reveal 转场,通用写法,适合 Shorts/Reels 的惊吓式 hook。" }, { "id": "hg-009", "model": "higgsfield-soul", "category": "technique-snippet", "title": "Higgsfield Genjutsu 运动迁移换角:主角+群演身份锁写法", "prompt": "CHARACTER AND BACKGROUND REPLACEMENT TASK\n\n@[Video 1]. Image 1 defines the lead's appearance, and Image 2 defines the supporting cast.\n\n1. MAIN CHARACTER:\nReplace ONLY the original lead with the man from @[Image 1]. Track this same person throughout the video, including close-ups, walking, turns, interactions and the finale. Preserve the source lead's screen position, action timing and apparent body scale. Image 1's full-body and facial views depict ONE man. Use his exact identity, body proportions and wardrobe; ignore the image background. He appears only once.\n\n2. SURROUNDING CHARACTERS:\nReplace EVERY OTHER PERSON with the complete masked red-jumpsuit character from @[Image 2], repeated as identical copies. Include front, side, middle and background rows, blurred figures and contact partners. Preserve the source crowd count, formation and individual positions. Leave no original supporting person unchanged. Ignore the image background.\n\n3. IDENTITY LOCK:\nImage 1 alone defines the lead's entire head and facial anatomy: facial width, eyes, nose, cheeks, jaw, hair, beard and the exact size, placement and angle of his ears. Do not inherit or blend the source actor's head geometry, face, tattoos or clothing.\nThe video supplies movement, head orientation, gaze and expression timing. Reproduce expression intent using the reference man's own facial structure, without reshaping his features.\nSupporting characters retain Image 2's complete mask and costume throughout. Masks remain rigid, with fixed painted expressions, and stay attached. No identity drift, costume blending, exposed crowd faces or duplicate leads.\n\n4. MOTION AND TIMING:\nPreserve the source duration, every cut and the original human choreography at its original timestamps. Each replacement follows the gestures, head turns, facing direction, walking speed and limb movements of the person it replaces. Preserve interactions with coherent hands and genuine shoulder contact. Do not add, remove or retime human actions.\n\n5. CAMERA AND FRAMING — VIDEO PRIORITY:\nVideo 1 alone defines camera angle, height, trajectory, lens perspective, framing, focus transitions and shot duration at every timestamp. Preserve the source pull-back and all other camera movements exactly as they occur; add none.\nKeep the lead's head and body the same apparent size and position within the frame as the source lead. Preserve supporting performers' relative screen scale, spacing, depth and occlusion.\nPreserve the source cropping of foreground bodies. Do not widen the shot to reveal full bodies, the complete crowd, more floor, the entire staircase or the vault. Do not add zoom, reframe, stabilize or change the source camera path. Maintain the source composition in 16:9.\n\n6. BANKNOTES:\nAdd only a few loose paper banknotes gently fluttering near the outer frame edges, with believable gravity and motion blur. Keep the lead's face, contact interactions and central route clear. No dense money rain. The vault door, architecture and staircase remain stationary in world space, showing only the perspective changes caused by the original camera.\n\nAVOID:\nAltered shot sizes, extra pull-back, smaller lead, wider crowd composition, additional floor revealed, Image 3 framing copied into the video, source-actor identity blending, exaggerated ears, warped faces, animated masks, original crowd remnants, extra or missing people, malformed hands, floating feet, flicker, excessive flying money, added scenes, text or logos.\n\n(作者注:示例里的 masked red-jumpsuit 角色和 banknotes 按自己角色/群演替换;若不换背景则删掉位置/灯光指令,改为保留原背景。)", "tags": [ "higgsfield-soul", "technique-snippet", "reference", "close-up", "rain", "crowd" ], "params": {}, "source": { "name": "beingovee (Substack)", "url": "https://beingovee.substack.com/p/viral-ai-video-prompt-25-ft-gener8ion", "tier": "reputable-3rd-party" }, "notes": "Higgsfield Genjutsu(Motion transfer)的新玩法:视频提供动作与机位,Image 1 换主角、Image 2 批量替换所有群演,含 identity lock / anti-drift / camera-priority 的完整防崩清单,是现有 hg- 条目没有的写法。" }, { "id": "so-021", "model": "sora-2", "category": "cinematic", "title": "孔雀岛延时巨变:从热带荒岛到璀璨都市", "prompt": "Theme: A lush tropical island shaped like a peacock undergoes rapid urbanization into a glittering megacity\n\nVisuals: Aerial view of a peacock-shaped island with natural vegetation forming tail-feather patterns, surrounded by turquoise ocean and blue sky. The island transforms through deforestation, construction, city-building, sunset, and finally a luminous night metropolis with a jewel-encrusted peacock head structure.\n\nCamera: Wide high-angle aerial drone view throughout, ending with a rapid zoom-in through the city streets to a low-angle close-up of the glowing peacock head.\n\nStyle: Photorealistic, cinematic time-lapse, vibrant tropical colors shifting to warm sunset tones and finally cool night illumination.\n\nAction + Sound Design:\n[0-2s] Wide aerial shot of the pristine green peacock-shaped island, vegetation forming detailed eye-spot tail patterns, turquoise water and bright blue sky — calm orchestral strings, gentle ocean waves cut\n[2-7s] The island rapidly transforms: green vegetation recedes into brown earth, logging operations begin, heavy machinery and cargo ships appear, construction cranes rise across the landscape — sounds of chainsaws, heavy machinery, building ambience cut\n[7-10s] Skyscrapers and city infrastructure rise rapidly across the island, roads and buildings take shape, the peacock head structure begins forming, daylight shifts to late afternoon — construction sounds intensifying, uplifting orchestral build cut\n[10-13s] The city is fully built as the sun sets, casting fiery orange and gold light across the water and buildings, the peacock tail pattern now defined by roads and districts, a cruise ship passes in the background — warm sunset orchestral swell, distant city hum cut\n[13-16s] Night falls, the city lights turn on illuminating the peacock tail pattern with golden streetlights, the diamond peacock head structure glows brilliantly against the dark sky — nocturnal orchestral score, gentle ocean waves, distant traffic cut\n[16-18s] The camera rapidly zooms forward from the wide aerial view, flying through illuminated skyscrapers and busy highways toward the glowing peacock head — rushing wind, building orchestral crescendo cut\n[18-20s] Low-angle close-up of the intricate jewel-encrusted peacock head structure glowing magnificently against the night sky, city lights and ocean in the background — powerful orchestral finale, shimmering magical tone", "tags": [ "sora-2", "cinematic", "close-up", "aerial", "drone", "night" ], "params": {}, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/726-viral-video-prompt-timelapse", "tier": "reputable-3rd-party" }, "notes": "9 月病毒社区爆款:20 秒分段时间轴 + 全程音效同步的延时转场写法(通用型,Ve o 3/Sora 2 均可);以异形岛屿形状做贯穿视觉锚点。" }, { "id": "ve-025", "model": "veo-3", "category": "creative", "title": "iPhone 叠塔上放火球熔化爆裂", "prompt": "(来源仅给出描述,按其给出的完整描述转录) First Frame Image Prompt: a man behind a white kitchen island with a tall stack of about 15 copper-colored smartphones. Static frontal medium shot, 9:16 vertical, photoreal product-demo look, cool white room.\n\nImage-to-Video Prompt: A glowing fiery orb is placed on top of the stack of phones. It melts through the phones layer by layer, then explodes into charred, blackened metal. Warm orb light against the cool white room. Timed sound cues: classical music bed, energy hum, sizzle, explosion, fading crackle across about 10-12 seconds.", "tags": [ "veo-3", "creative", "product", "explosion" ], "params": { "aspect": "9:16", "duration": "12s" }, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/10-top-viral-ai-tools-and-tips-today-6b4", "tier": "reputable-3rd-party" }, "notes": "首帧图 + I2V 的产品 VFX 演示爆款写法(通用型):铜色手机叠塔被火球逐层熔穿后爆成焦黑金属,带 10-12 秒定时的音效分层。" }, { "id": "ve-026", "model": "veo-3", "category": "comedy", "title": "泳池派对微缩滑板追逐(一镜到底)", "prompt": "Use Image 1 only for the tiny skateboarder identity: exact face, hairstyle, skin tone, body build, and appearance. Preserve identity 100%. Adult male only. Stylish youthful streetwear: light bomber jacket or fitted overshirt, plain T-shirt, slim jeans, clean sneakers (waterproof sheen acceptable). Use Image 2 only for the giant child's appearance and behavior. Do not mix identities.\n\nCreate a 15-second cinematic miniature chase around a giant backyard pool deck during a kids' pool party. ONE continuous shot, NO cuts. A tiny real adult man rides a miniature skateboard across sun-warmed pool deck tiles while a giant child plays in and around the pool. He is a real human shrunk to miniature size, not a toy, doll, cartoon, or child. The child, pool floats, water splashes, and deck furniture feel enormous.\n\nCamera: ultra-low deck-level chase camera at skateboard height, tightly behind or slightly beside the rider. It moves forward along the pool deck path from start to finish. Never reset, jump, teleport, change scene, cut, hide cuts, montage, or transition. Camera accelerates with the skateboard, leans with turns, dips during dodges, and reacts to near-collisions.\n\nEnvironment: bright sunny poolside, warm textured pool deck tiles with wet reflective patches. Giant pool noodles, a beach ball, stacked towels, flip-flops, a cooler, scattered water puddles, pool toys near the edge. Cheerful, sunny, vibrant summer atmosphere. Keep one narrow continuous deck path.\n\nOnly ONE child exists as the active giant figure. No duplicate child, second kid, extra hands, or extra limbs. Any hand, arm, or partial body belongs to the same single child. The child stays in one consistent area ahead and slightly beside the path, never teleporting. Joyful and energetic, not violent.\n\nAction is one connected chain reaction. The skateboarder never stops. Start low behind him racing along the sunlit deck between giant pool toys. The child does a cannonball jump nearby, sending a giant wave of splashed water sliding across the deck path; he swerves through the spray by inches. The same child's momentum knocks a beach ball onto the dock, bouncing into the path. He carves tightly around the bouncing ball. The child kicks a floating pool noodle up onto the deck, it slides and blocks part of the path. He squeezes through the narrow gap beside it, wheels skimming a wet patch. The child reaches down for one final splash toward the tiny figure, giant hand and water droplets flying. He crouches low and slips beneath a stack of pool towels on a lounge chair by inches. Camera follows into a shaded gap behind the pool equipment shed. End with a split-second escape feeling.\n\nAudio: urgent cinematic tension music, fast suspense rhythm, summer fun energy. Sounds: skateboard wheels on wet tile, water splashing, ball bouncing, child laughter and movement, pool noodle slide, sharp carving turns.\n\nImportant: keep protagonist tiny, realistic, adult male, identity-locked. Keep child huge and singular. Maintain consistent scale, layout, and physical continuity. Preserve face whenever visible. Realistic cinematic lighting, shallow depth of field, natural motion blur, detailed textures, bright sunny grading, 8K realism.\n\nNegative prompt: cartoon, doll, toy person, plastic skin, female protagonist, adult-only crowd, changed face, wrong identity, duplicate child, second kid, extra hands, extra limbs, duplicated parts, teleporting child, cuts, montage, camera reset, inconsistent layout", "tags": [ "veo-3", "comedy", "crowd" ], "params": {}, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/735-viral-video-prompt-miniature", "tier": "reputable-3rd-party" }, "notes": "9 月社区最强身份锁写法之一:Image 1 锁微缩骑手、Image 2 锁巨人小孩,15 秒无剪辑连续追逐 + 连锁反应动作链 + 详尽 negative prompt(通用型,模型无关)。" }, { "id": "so-022", "model": "sora-2", "category": "creative", "title": "打破画框:从复古电视走进现实,再踏入黑边", "prompt": "First Frame Image Prompt @Image 1\n\nA wide cinematic shot of a cozy, retro 1970s living room at night, warmly lit by a table lamp with a soft orange shade on the right and a stone fireplace with a glowing fire on the left. The room has earthy tones: warm brown wooden floor and furniture, a muted green mid-century armchair, built-in bookshelves filled with books behind it, and a patterned rug. On the right, a large vintage wooden television set sits on a low wooden console table, next to a rotary phone on a shelf. Inside the TV screen, an Asian man in his 30s, wearing a simple, timeless outfit (a button-up shirt and trousers), stands against a flat, light gray background, looking directly ahead with a calm, composed expression. Thick black cinematic letterbox bars frame the top and bottom of the image. The atmosphere is nostalgic, surreal, and warmly atmospheric.\n\nImage-to-Video Prompt\n\nTheme: A man breaks out of a vintage TV screen, crosses into the physical living room, and ultimately steps beyond the video frame itself into the black letterbox space\n\nVisuals: Cozy 1970s living room with warm lamp and fireplace lighting, vintage wooden TV, green armchair, bookshelves, patterned rug — contrasted against the sterile gray broadcast world inside the television. An Asian man with a calm, confident presence transitions from the flat TV image into a three-dimensional physical presence, with lighting realistically adapting from bright flat broadcast light to warm ambient room light\n\nCamera: Static wide shot holding the room, then a slow, smooth dolly zoom-in toward the man as they walk forward, tilting upward as they climb out of frame, ending on the man standing fully within the black bar space, facing the viewer\n\nStyle: Cinematic, magical realism, retro-futuristic, warm earthy color palette, soft atmospheric lighting, surreal fourth-wall break\n\nVO Voice Style: Calm, warm, slightly contemplative male voice — measured pacing, cinematic trailer tone, with a sense of quiet wonder building into inspiration\n\nAction + Sound Design:\n[0-2s] The man stands inside the vintage TV screen while the room remains still. Subtle electronic static hum and warm fireplace ambience.\nVO: \"We spend our whole lives... watching.\"\nOn-screen text appears: \"Who said you have to stay inside the lines?\"\n[2-7.5s] The man reaches forward, physically steps out through the TV screen into the living room, and begins walking calmly toward the camera. The lighting on his clothing and skin dynamically shifts from flat broadcast light to warm room light. Soft footsteps on wood, subtle electronic glitch sound as he crosses the threshold.\nVO: \"But somewhere along the way... we forgot we could step in.\" @Image 2\nOn-screen text appears: \"Break the frame.\"\n[7.5-10.5s] The man continues walking toward the camera, climb over the sofa and stepping directly into the black letterbox bar at the bottom of the frame. His full body — legs, torso, and head — crosses into the black space, now standing there as if it's solid ground.\nVO: \"The frame was never the limit.\"\nOn-screen text appears: \"Step outside.\"\n[10.5-12s] Now standing fully inside the black bar space, facing the viewer, the man raises his right hand and points his index finger directly at the camera, with a subtle, confident smile.\nVO: \"So— what's stopping you?\"\nFinal on-screen text appears: \"Make your own!\"", "tags": [ "sora-2", "creative", "night" ], "params": {}, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/743-viral-video-prompt", "tier": "reputable-3rd-party" }, "notes": "第四面墙破框创意:首帧图 + I2V + VO + 屏幕文字的完整组合写法(通用型);灯光从“电视平光”动态切换为“房间暖光”是关键细节。" }, { "id": "ve-027", "model": "veo-3", "category": "product-commercial", "title": "奢华香水广告 10 秒(开盖→喷雾→闻香→产品英雄镜)", "prompt": "Cinematic luxury perfume commercial, 10 seconds. A handsome, confident man with slicked-back dark hair @Image 1 and a light stubble, wearing a tailored black blazer over a black fitted shirt with the top button open, stands against a soft pale blue-grey studio background. He holds a glass bottle of Sifu Yik Eau de Parfum @Image 2 with both hands toward the camera, presenting it. Close-up of his strong hands removing the black cap, revealing the silver spray nozzle. He sprays a fine mist onto his inner wrist, with his sleeve cuff slightly pulled back, in a bright window-lit setting. He then lifts his wrist to his face, closes his eyes, and inhales the scent with a subtle satisfied smile. Final shot: the bottle fills the frame in sharp focus with the Sifu Yik logo clearly visible, while he is softly blurred behind it. Soft natural lighting, shallow depth of field, high-end fashion film aesthetic, slow elegant movements, premium commercial look.\nVoiceover: \"Getting ready, one final touch. Sifu Yik, make it yours.\"", "tags": [ "veo-3", "product-commercial", "close-up" ], "params": { "duration": "10s" }, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/746-viral-video-prompt-perfume-ad", "tier": "reputable-3rd-party" }, "notes": "AI 香水广告的完整分镜写法:开盖/喷雾/闻香/英雄镜四镜一气呵成 + voiceover;@Image 1/2 分别锁人物与产品身份(通用型)。\n演示视频:https://sifuyik.substack.com/api/v1/video/upload/a66b7825-e192-4315-b3ad-e8df50c300b7/src?override_publication_id=7223942&preview=false&type=hls" }, { "id": "so-023", "model": "sora-2", "category": "image-to-video", "title": "古地图活化成微缩世界(4 段定时)", "prompt": "(来源仅给出描述,按其给出的完整描述转录) #745 Viral Video Prompt: Map to Live — a copy-paste vertical 9:16 4K HDR video prompt where an antique paper map physically transforms into a living miniature world. 4 timed scenes: 0 to 3 seconds macro hook of the ordinary map, 3 to 7 transformation, 7 to 11 dive into the world, 11 to 15 aerial reveal. Mountains rise, rivers carve and flow, forests grow, roads emerge and miniature towns build along them. A tiny vintage car drives through a mountain village with glowing windows and pedestrians. Final frame pulls up to reveal the whole living continent still sitting on the paper map. Strict consistency rules: no cuts, no teleporting, no cartoon look, the map stays recognizable throughout.", "tags": [ "sora-2", "image-to-video", "aerial" ], "params": { "aspect": "9:16", "duration": "3s" }, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/15-top-viral-ai-tools-and-tips-today-257", "tier": "reputable-3rd-party" }, "notes": "“等待-发生了什么”式病毒结构:古董纸地图 4 段定时转场成活微缩大陆,末帧拉回证明地图仍在;严格无剪辑一致性规则(通用型)。\n演示视频:https://sifuyik.substack.com/api/v1/video/upload/06424e34-e19b-456d-b21e-3bb31fb6d923/src?override_publication_id=7223942&preview=false&type=hls" }, { "id": "w2-001", "model": "wan-2.6", "category": "dialogue-driven", "title": "Wan 2.6 R2V 多 reference 对白表演", "prompt": "Video 1 holds Image 3 and plays a soothing country folk song on the chair from Image 4, saying: \"The weather is so nice today.\" Image 1, holding Image 2, walks past Video 1, places Image 2 on the table next to Video 1, and says: \"That sounds beautiful. Can you play it again?\"", "tags": [ "wan-2.6", "dialogue-driven" ], "params": {}, "source": { "name": "阿里云文档", "url": "https://help.aliyun.com/en/model-studio/video-to-video-guide", "tier": "official" }, "notes": "阿里云官方 Wan-R2V 文档示例:视频+图片+音频多 reference 联合驱动的多人对白表演,展示 Wan 2.6 代 reference-to-video 新范式\n演示视频:https://help-static-aliyun-doc.aliyuncs.com/file-manage-files/en-US/20260416/mncxqr/Wan2.7_Video_Reference_result.mp4" }, { "id": "ra-001", "model": "runway-aleph-2.0", "category": "video-edit", "title": "Runway Aleph 2.0 锚点品牌替换", "prompt": "Apply the brand identity shown in the reference image to the plain white sneaker. Keep the sneaker shape, the camera orbit, the matte surface, lighting, and framing exactly as in the source.", "tags": [ "runway-aleph-2.0", "video-edit", "reference" ], "params": {}, "source": { "name": "runware/runware-skills", "url": "https://github.com/runware/runware-skills/blob/HEAD/skills/edit-video/references/examples.md", "tier": "reputable-3rd-party" }, "notes": "Aleph 2.0 anchor-frame 精确编辑范式:prompt 只命名变化、reference 首帧图定义精确外观,解决品牌级精确替换" }, { "id": "ra-002", "model": "runway-aleph-2.0", "category": "technique-snippet", "title": "Aleph 2.0 外科手术式单点编辑", "prompt": "Replace the gray sky with a blue sky with clouds", "tags": [ "runway-aleph-2.0", "technique-snippet" ], "params": {}, "source": { "name": "explainx.ai", "url": "https://explainx.ai/blog/runway-aleph-2-vs-gemini-omni-video-2026", "tier": "unverified-3rd-party" }, "notes": "Aleph 2.0 “surgical edit”一句话范式代表:只改一处、其余保持,区别于全片重绘的 remix 路线" }, { "id": "gm-011", "model": "gemini-omni", "category": "video-edit", "title": "Gemini Omni 一句话转绘", "prompt": "Turn this cooking video into a Studio Ghibli animation", "tags": [ "gemini-omni", "video-edit" ], "params": {}, "source": { "name": "explainx.ai", "url": "https://explainx.ai/blog/runway-aleph-2-vs-gemini-omni-video-2026", "tier": "unverified-3rd-party" }, "notes": "Gemini Omni “remix”重绘范式代表:一句话整体转绘,与 Aleph 精确编辑形成对照" }, { "id": "lu-001", "model": "luma-ray-3.2", "category": "quiet-moments", "title": "Luma Ray 3.2 观察式短片公式", "prompt": "Locked medium-wide shot of a red folding bicycle beside a rain-streaked café window. A barista enters from the left, closes the bicycle and carries it out of frame. Soft overcast daylight, realistic metal reflections, quiet observational film style. Keep the bicycle frame geometry and red color consistent.", "tags": [ "luma-ray-3.2", "quiet-moments", "rain" ], "params": {}, "source": { "name": "magichour.ai", "url": "https://magichour.ai/blog/how-to-use-luma-ai-dream-machine", "tier": "unverified-3rd-party" }, "notes": "Ray 3.2 官方 shot-direction 公式示范(机位+主体+单一动作+场景+光影+风格+一致性约束),9-13 发布的 3.2 新模型" }, { "id": "lu-002", "model": "luma-ray-3.2", "category": "video-edit", "title": "Luma Ray 3.2 表演迁移换装", "prompt": "Keep the performer's timing and body motion. Change the jacket to a dark red raincoat and the setting to a quiet train platform.", "tags": [ "luma-ray-3.2", "video-edit", "rain" ], "params": {}, "source": { "name": "magichour.ai", "url": "https://magichour.ai/blog/luma-dream-machine-review", "tier": "unverified-3rd-party" }, "notes": "Ray 3.2 Modify 范式:保留原片动作时序、只换服装与场景,展示 20 秒编辑能力" }, { "id": "gm-012", "model": "gemini-omni", "category": "nature", "title": "Gemini Omni Vids 航拍雪山", "prompt": "A cinematic aerial view of snow-covered mountains at sunrise, with clouds moving slowly through the valley and warm golden lighting.", "tags": [ "gemini-omni", "nature", "aerial" ], "params": {}, "source": { "name": "aitoolharbor.blogspot.com", "url": "http://aitoolharbor.blogspot.com/2026/09/google-vids-ai-video-generator-2026.html", "tier": "unknown" }, "notes": "Google Vids 9-23 免费开放 Gemini Omni 1.1 Flash 后的官方写法示例,配合 exact duration 与 scene extension 新功能" }, { "id": "ve-028", "model": "veo-3", "category": "image-to-video", "title": "Instagram 帖子互动:人物打破 feed 边界摸头杀", "prompt": "First-Frame Image Prompt:\n\nCreate a photorealistic Instagram feed in exact 4:5 aspect ratio using the uploaded images as visual references. TOP POST. Use [TOP IMAGE]. Show [TOP CHARACTER] performing [TOP ACTION]. The subject must extend naturally beyond the bottom edge and interact with the bottom post.\n\nBOTTOM POST: Use [BOTTOM IMAGE]. Show [BOTTOM CHARACTER/SUBJECT] performing [BOTTOM ACTION]. The top subject must physically pass through the feed boundary and continue into the bottom image with realistic perspective, scale, lighting, depth, occlusion, and shadows.\n\nINSTAGRAM UI: Recreate a real Instagram dark-mode feed with accurate layout, typography, spacing, icons, alignment, and hierarchy. Engagement row: [LIKES] / [COMMENTS] / [SHARES]. Then: [LIKE ICON] Liked by [LIKED-BY NAME]. Caption: [USERNAME] [ONE-LINE CAPTION] [EMOJI]. Below: [PROFILE PICTURE] [USERNAME] (use the uploaded profile picture as the avatar).\n\nDESIGN & STRUCTURE:\n- Dark/black Instagram UI.\n- Two vertically stacked posts with no horizontal divider.\n- Square image corners; no rounded corners.\n- Both images extend fully from left to right edges with no side margins.\n- Keep the dark feed area between posts where the UI appears.\n- Top post: no author profile picture or username; show image, engagement row, liked-by, and caption only.\n- Bottom post: show only [PROFILE PICTURE], [USERNAME], and three-dot menu; no engagement row, liked-by, or caption.\n- The overlapping top subject must pass through the black UI area and appear physically in front of it.\n- Add realistic cast/contact shadows following the scene's lighting.\n- Preserve uploaded images, colors, composition, and important details.\n- Keep all text/icons sharp, aligned, and legible.\n- No unnecessary borders, distortions, extra UI, or random text.\n- Do not redesign or reinterpret the Instagram interface.\n- Make the interaction seamless and physically realistic, not like a graphic mockup.\n- Use the uploaded reference images for the exact feed composition and interaction style. Aspect ratio: 4:5\n\nImage-to-Video Prompt:\n\nStyle: Warm, playful, lighthearted social-media-style comedy. Soft studio lighting, neutral gray backdrop, smooth natural motion. The camera stays static, like a phone screen recording of an Instagram feed.\n\nFirst frame: Two stacked Instagram-style posts. The man in the white polo is in the top post, with his arm reaching out of the frame. His hand rests on the head of the woman in the bottom post. She has a small, smug smile.\n\nTimeline:\n0-2s: The man smiles warmly at the camera and gives the woman's head a few gentle, affectionate pats. The woman keeps her closed-lip smile, but her eyes slide sideways toward his hand. The like count ticks up from 8,432 to 8,433.\n2-4s: The woman's smile flattens into a deadpan look. She slowly turns her head and gives his hand a long, unimpressed side-eye. He keeps smiling and pats a little faster, pretending not to notice.\n4-6s: Fed up, she tilts her head up and blows a stray strand of hair off her face. The man's grin turns sheepish and he lifts his hand an inch. She quickly tilts her head back under it, as if she secretly wants the pats to continue.\n6-8s: The man laughs and ruffles her hair a little, and she scrunches her nose and tries not to smile. The caption \"Lucky to have you\" pulses softly. The comment count rises from 214 to 215.\n8-10s: The woman finally breaks into a big, genuine smile and leans into his hand. The man looks down at her with affection. The heart icon pops with a tiny animation, and the frame holds on the two of them.\n\nAudio: Light, cheerful ukulele or soft acoustic music. Subtle UI \"pop\" sounds when the likes and comments tick up. A soft chuckle from the man around 6s, and a quiet giggle from the woman at 8s. No dialogue.\n\nNegative prompt: No camera shake, no face morphing, no extra fingers, no text distortion, no sudden lighting changes, no changes to clothing or hairstyle.\n\n(Use the reference image as first frame.)", "tags": [ "veo-3", "image-to-video", "dialogue", "reference", "comedy" ], "params": { "aspect": "4:5" }, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/751-viral-video-prompt-instagram", "tier": "reputable-3rd-party" }, "notes": "打破第四面墙的病毒格式:上方帖子人物伸手穿过 UI 摸下方人物头,点赞数实时跳动;首帧图+I2V 双 prompt 结构完整,Veo 3 文字渲染能力正好对口\n(来源为模型无关 prompt,模型归属为编辑指派;日期为相对时间推算)" }, { "id": "ve-029", "model": "veo-3", "category": "image-to-video", "title": "三招食材新鲜度测试:三文鱼/整鱼/西兰花 30 秒教程", "prompt": "First Frame Image Prompt (<<>>):\n\nA medium close-up shot of a man, wearing a white distressed t-shirt with a USA flag graphic, standing in a bright sunroom or enclosed patio with large glass windows behind him showing a rainy green backyard. He holds a large raw salmon fillet over a clear rectangular glass tank filled with hot water on a marble countertop. The man looks directly at the camera with an instructional expression. Natural daylight, bright and clean, shallow depth of field, smartphone vertical video aesthetic, 720x1280 aspect ratio.\n\nImage-to-Video Prompt:\n\nTheme: Food quality and safety tutorial - three quick kitchen tests for salmon, whole fish, and broccoli\n\nVisuals: Bright sunroom with large glass windows, rainy green outdoor background, marble countertop, clear glass tanks and bowls, raw salmon fillet, whole silver fish, raw broccoli florets, wooden cutting board with raw chicken, salmon, and trout. Prominent yellow and white bold text overlays summarizing each tip. Natural daylight, clean and slightly overexposed smartphone look.\n\nCamera: Static medium close-up on the presenter and countertop for the entire video, with occasional slow digital zooms into the food reactions and close-ups of the water tank and broccoli bowl.\n\nStyle: Bright, informative, DIY social-media tutorial aesthetic; natural color grading with emphasis on orange salmon and green broccoli; bold yellow text overlays; vertical smartphone video format.\n\nAction + Sound Design:\n[0-8s] The man holds up a raw salmon fillet and lowers it into the clear tank of hot water. The salmon quickly begins to flake and break apart into small pieces in the water. He gestures toward the tank and speaks directly to camera. On-screen text appears: \"HOT WATER\" and \"IF IT FALLS APART EASILY THE MEAT ISN'T REAL.\" Sound: casual spoken narration + light water movement + soft ambient room tone\n[8-15s] He removes the salmon and places a whole silver fish into a clear tank of cold water. The fish sinks slowly to the bottom. He points at the tank while explaining the freshness test. On-screen text appears: \"COLD WATER,\" \"IF IT SINKS,\" and \"IF IT FLOATS IT'S BEEN SITTING ON THE SHELF.\" Sound: continued narration + gentle water splash + subtle tank tapping\n[15-23s] He places a glass bowl of raw broccoli florets on the counter and pours water over them, then adds a splash of white vinegar from a clear bottle. The camera slowly zooms in to reveal tiny white worm-like insects and eggs floating in the liquid. On-screen text appears: \"WATER WITH A SPLASH OF VINEGAR,\" \"ANY HIDDEN INSECTS AND THEIR EGGS WILL COME LOOSE IN THE WATER.\" Sound: narration + pouring water + light splashing + subtle vinegar bottle cap sound\n[23-30s] The man holds up a wooden cutting board displaying a whole raw chicken, several salmon pieces, and a whole fish. He smiles slightly and gestures toward the camera, encouraging viewers to save and follow. The camera slowly zooms in on the cutting board. On-screen text appears: \"SAVE THIS AND COMMENT WITH OTHER FOODS,\" \"FOR MORE FOLLOW,\" and \"DON'T MISS OUT.\" Sound: upbeat closing narration + light cutting board tap + ambient room tone", "tags": [ "veo-3", "image-to-video", "close-up", "rain", "food", "tutorial" ], "params": {}, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/752-viral-video-prompt-food-tips", "tier": "reputable-3rd-party" }, "notes": "30 秒竖屏教程类病毒模板:三段式食物测试+黄色大字报 overlay,手机实拍美学,科普/带货号可直接套用改食材\n(来源为模型无关 prompt,模型归属为编辑指派;日期为相对时间推算)" }, { "id": "so-024", "model": "sora-2", "category": "cinematic", "title": "名人夜晚离场:旋转门签名+跑车疾驰 10 秒", "prompt": "Image-to-Video Prompt:\n\nTheme: A celebrity named \"Sifu\" exits a building at night, signs an autograph for fans behind a barricade, then makes a quick getaway in a sports car\n\nVisuals: Glass revolving door entrance with warm interior light spilling onto a wet dark pavement, a man in a grey shirt and dark trousers walking with two security guards in black suits, a crowd of fans and photographers behind a metal barrier holding up phones and cameras, camera flashes popping, velvet rope stanchions in the foreground, a sleek sports car waiting at the curb\n\nCamera: Handheld documentary/paparazzi style, eye-level medium shot, camera tracks gently forward and follows the man as he walks toward the lens and stops to sign, then pans/follows as he moves to the car and drives off\n\nStyle: Realistic, cinematic nighttime street photography, high contrast between bright building interior and dark exterior, dynamic flash lighting, glossy wet ground reflections, energetic celebrity-arrival atmosphere\n\nAction + Sound Design:\n[0-1.5s] The man and two security guards emerge from the revolving door into the night, the crowd behind the barricade visible and cheering \"Sifu\", camera shutters clicking\n[1.5-4s] He walks toward the camera with a smile, waving to the fans, while camera flashes pop around him and the guards stay close on either side\n[4-7s] He stops, takes a pen, and signs an autograph on a white notebook held out from the lower left of frame, the guards waiting patiently beside him\n[7-8.5s] He finishes signing, hands the pen back, waves once more to the crowd, then turns and strides confidently toward the waiting sports car\n[8.5-10s] He swings open the door, hops into the driver's seat in one fluid motion, the engine roars to life, and the car peels away from the curb into the night, tail lights streaking across the wet pavement as the crowd reacts with surprise and excitement", "tags": [ "sora-2", "cinematic", "night", "crowd" ], "params": {}, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/753-viral-video-prompt-celebrity", "tier": "reputable-3rd-party" }, "notes": "狗仔跟拍式现实主义短片模板:旋转门→签名→跑车离场一镜式调度,湿地面反光+闪光灯氛围,10 秒完整叙事弧,换脸即用\n(来源为模型无关 prompt,模型归属为编辑指派;日期为相对时间推算)" }, { "id": "ve-030", "model": "veo-3", "category": "image-to-video", "title": "Wants vs Needs:红绿双柱信息图励志口播", "prompt": "First Frame Image Prompt:\n\nThe man sits at a dark wooden table against a plain solid light gray background. He looks directly at the camera with an open, engaging expression, hands resting on the table. Above his head is a clean infographic layout: on the left, bold red text reads \"YOU WANT\" above a vertical stack of five rounded red rectangular bars, with the top bar labeled \"money\" in white text; on the right, bold green text reads \"YOU NEED\" above a vertical stack of five rounded green rectangular bars, with the top bar labeled \"skills\" in white text. The lighting is bright, even, and professional, like a studio talking-head setup. The composition is vertical 9:16, clean and minimal, suitable for short-form social media content.\n\nImage-to-Video Prompt:\n\nTheme: Motivational \"wants vs needs\" self-help breakdown - a direct-to-camera speaker explains the real price behind common desires.\n\nVisuals: A clean talking-head setup with a middle-aged man in a light blue t-shirt sitting at a dark wooden table against a solid light gray wall. Above his head, two infographic columns display stacked colored bars: red \"YOU WANT\" on the left, green \"YOU NEED\" on the right. Text labels appear one by one on the bars as he speaks.\n\nCamera: Static medium close-up shot, chest-up framing, vertical 9:16, centered on the speaker, sharp focus on his face with the graphic overlay occupying the upper portion of the frame.\n\nStyle: Bright, clean, professional studio lighting, modern short-form educational/motivational content aesthetic, minimal background, bold typography.\n\nAction + Sound Design:\nThe man begins speaking directly to camera with calm, confident energy - clean studio audio, no background music\n[0-3s] The top red bar under \"YOU WANT\" is labeled \"money\" and the top green bar under \"YOU NEED\" is labeled \"skills\" as he says: \"If you want money, you have to have valuable skills.\"\n[3-6s] The second red bar is labeled \"success\" and the second green bar is labeled \"risk\" as he says: \"If you want success, you have to take risk.\"\n[6-9s] The third red bar is labeled \"respect\" and the third green bar is labeled \"lead\" as he says: \"If you want respect, you're gonna have to lead by example.\"\n[9-13s] The fourth red bar is labeled \"time\" and the fourth green bar is labeled \"no control\" as he says: \"If you want more time, this one's gonna hurt, you're gonna have to let go of control.\"\n[13-16s] The bottom red bar is labeled \"grow\" and the bottom green bar is labeled \"let go\" as he says: \"If you wanna grow your future, you're gonna have to let go of your past.\"\n[16-19s] A bright warm flash transition wipes away all graphics, leaving only the man at the table, and on-screen captions appear as he says: \"And if you wanna learn how, follow me for more.\" - friendly, inviting delivery\n\n(Use the reference image as first frame.)", "tags": [ "veo-3", "image-to-video", "reference", "close-up" ], "params": { "aspect": "9:16" }, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/755-viral-video-prompt-motivational", "tier": "reputable-3rd-party" }, "notes": "对口型+信息图叠加的标准励志号模板:红绿双柱随台词逐条点亮,首帧图锁定排版,成长/知识号改词即复用\n(来源为模型无关 prompt,模型归属为编辑指派;日期为相对时间推算)" }, { "id": "so-025", "model": "sora-2", "category": "image-to-video", "title": "人类 vs 银背大猩猩:竞技场扇耳光大赛", "prompt": "First-Frame Image Prompt:\n\nA wide, eye-level shot of a brightly lit indoor sports arena. In the center, a slap-fighting competition is about to begin. On the left stands a muscular man wearing a tight black t-shirt with \"SLAP\" written in white, black sweatpants, and red athletic shoes. On the right stands a massive, realistic silverback gorilla standing upright. Between them is a black cylindrical podium with neon green and pink accents and the text \"10X.\" In the background, between the two competitors, stands a referee in a black polo shirt. Behind them, a large, slightly blurred crowd sits in stadium seating under bright, dramatic overhead lighting. Hyper-realistic, high-definition sports photography, cinematic arena atmosphere.\n\nImage-to-Video Prompt:\n\nTheme: Surreal human vs gorilla slap-fighting showdown in a professional sports arena\n\nVisuals: A packed indoor arena under dramatic overhead lighting, a black cylindrical slap podium with neon green/pink accents and \"10X\" branding, a muscular man in a tight black \"SLAP\" t-shirt and red shoes facing an enormous upright silverback gorilla, with a referee standing between them\n\nCamera: Wide establishing shot -> medium close-up on impact -> slow-motion close-up for the final strike\n\nStyle: Hyper-realistic sports photography, high-intensity, dark arena atmosphere with spotlighted action, exaggerated impact physics, cinematic slow motion\n\nAction + Sound Design:\n[0-2.5s] Wide establishing shot of the arena. The referee raises a hand and shouts \"Fight!\" as the man and gorilla face off across the podium - ambient crowd noise, tense atmosphere [cut]\n[2.5-5s] Medium close-up of the man. The gorilla's massive hand swings in from the right and slaps the man's left cheek with extreme force, sending a cloud of white powder and sweat erupting from the impact - loud exaggerated smack sound + crowd gasp [cut]\n[5-7.5s] The man absorbs the blow, his face rippling but staying upright. He clenches his jaw, glares back with determination, and winds up his right arm - heavy breathing + dramatic low-frequency rumble [cut]\n[7.5-10s] Slow-motion close-up profile shot. The man's muscular arm swings forward and his open palm connects solidly with the side of the gorilla's face, sending a shockwave through the gorilla's fur and another puff of white powder into the air - heavy impact thud + deep bass drop", "tags": [ "sora-2", "image-to-video", "close-up", "slow-motion", "neon", "crowd" ], "params": {}, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/756-viral-video-prompt-slap-competition", "tier": "reputable-3rd-party" }, "notes": "荒诞对决类病毒喜剧:扇耳光竞技的慢动作冲击物理+粉尘爆发,四段式剪辑节奏,强视觉奇观易出爆款\n(来源为模型无关 prompt,模型归属为编辑指派;日期为相对时间推算)" }, { "id": "so-026", "model": "sora-2", "category": "comedy", "title": "倒霉青蛙 30 秒:珍珠→河狸→毛毛虫连环喜剧", "prompt": "Reference Images (generate first):\n\nFrog Character Sheet:\nCharacter reference sheet of the green frog from the unlucky frog slapstick comedy. 3D Pixar/Dreamworks animation style, clean white background, multiple poses and expressions in a grid layout. Show front view, side view, 3/4 view, curious expression reaching for a pearl, nervous smile, pain expression with a swollen red finger, dazed expression after hitting a tree, and final close-up with comically enormous swollen bright red lips. Small plump green body, short limbs, large bulbous expressive eyes, wet glossy skin with lighter green highlights, wide mouth. High quality, detailed textures, vibrant colors, soft natural lighting.\n\nOtter Character Sheet:\nCharacter reference sheet of the grumpy brown otter with beaver-like features from the unlucky frog slapstick comedy. 3D Pixar/Dreamworks animation style, clean white background, multiple poses and expressions in a grid layout. Show front view, side view, 3/4 view, angry glare, chasing pose, tail-swinging batting pose, and with a clam balanced on its head. Stocky brown furry body, short brown fur, grumpy face with narrowed eyes and visible buck teeth, broad flat tail. High quality, detailed fur texture, vibrant colors, soft natural lighting.\n\nCaterpillar Character Sheet:\nCharacter reference sheet of the colorful spiky caterpillar from the unlucky frog slapstick comedy. 3D Pixar/Dreamworks animation style, clean white background, multiple views and poses in a grid layout. Show top view, side view, front view, curled pose, and stretched pose. Small segmented cylindrical body with bright multicolored segments in green, yellow, red, blue and orange, covered in sharp spikes and bristles. High quality, detailed texture, vibrant colors, soft natural lighting.\n\nVideo Prompt:\n\nTheme: Unlucky green frog <<>> slapstick comedy in a sunny pond\n\nVisuals: A small cartoon green frog with large expressive eyes, wet glossy skin, and a plump body; a grumpy brown beaver with buck teeth, a broad flat tail, and a clam balanced on its head; a tiny multicolored spiky caterpillar with green, yellow, red, orange, and blue segments; a serene pond with lily pads, tall reeds, a muddy bank, and a large tree trunk\n\nCamera: Wide establishing shot at the start, close-ups on facial reactions, quick cuts during the chase, a long tracking shot as the frog flies through the air, and a final extreme close-up on the swollen lips\n\nStyle: 3D Pixar/Dreamworks feature animation, vibrant saturated colors, realistic textures on stylized characters, soft warm natural outdoor lighting, exaggerated cartoon physics, slapstick comedy timing\n\nAction + Sound Design:\n[0-2s] Wide establishing shot of a sunny pond. The green frog sits on the muddy bank, eyes wide and curious, slowly reaching toward a large glowing pink pearl inside an open clam shell. Soft ambient pond sounds: gentle water lapping, distant birds, light wind through reeds.\n[2-3s] Close-up on the frog's face: pupils dilated, tongue slightly out, finger hovering just above the pearl. Tense silence with a faint magical shimmer from the pearl.\n[3-4s] The clam snaps shut on the frog's finger with a sharp CLACK. The frog's eyes bulge, mouth drops open in shock. Loud snap sound.\n[4-6s] The frog yanks its hand back and hops backward, clutching its throbbing swollen red finger. Pained \"YOW!\" vocalization, rapid heartbeat sound.\n[6-8s] The frog stumbles and accidentally kicks the clam shell, sending it spinning into the air. Whoosh sound, panicked breathing.\n[8-9s] The clam arcs through the air and lands squarely on the head of a grumpy brown beaver <<>> sitting nearby. Dull thud, beaver's surprised grunt.\n[9-11s] The beaver slowly turns its head toward the frog, eyes narrowing, buck teeth gritted, the clam still balanced on its head. Low ominous rumble.\n[11-13s] The frog looks up, frozen in fear, then forces a nervous toothy smile with a visible sweat drop. Awkward gulp, nervous giggle.\n[13-14s] The frog's smile drops; it spins around and sprints away in panic. Fast scurrying footsteps, splashing mud.\n[14-16s] The beaver snarls and charges after the frog on all fours, kicking up dust from the muddy bank. Heavy charging footsteps, angry growl.\n[16-18s] The frog leaps into the air trying to escape; the beaver pivots and swings its broad flat tail like a baseball bat. Tail swoosh, frog's scream.\n[18-19s] The frog is struck mid-air and spins wildly out of control. Impact smack, whoosh.\n[19-21s] Wide tracking shot: the frog flies across the pond in a high arc, limbs flailing, heading toward a large tree trunk on the far bank. Wind rushing sound.\n[21-22s] The frog splats flat against the tree trunk with a wet SPLAT, then peels off and slides down into the mud. Squish, mud splatter.\n[22-23s] The frog sits dazed in the mud, stars circling its head, tongue lolling out. Dizzy sound effect, tweeting birds.\n[23-25s] The frog's eyes refocus and spot a small bright multicolored spiky caterpillar <<>> crawling on a nearby leaf. Curious \"hmm\" sound, leaf rustle.\n[25-27s] The frog's tongue shoots out fast, wraps around the caterpillar, and snaps back into its mouth. Tongue snap, slurp.\n[27-28s] The frog chews once, eyes widen in pain, and spits the caterpillar out in a puff of dust and spikes. Crunch, \"OW!\", spitting sound.\n[28-30s] Extreme close-up on the frog's face: its lips swell to enormous, glossy bright red proportions. Comic rubber-stretch swelling sound with a pop, ending on a sad groan.", "tags": [ "sora-2", "comedy", "reference", "close-up", "comedy" ], "params": {}, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/757-viral-video-prompt-frog-adventure", "tier": "reputable-3rd-party" }, "notes": "三角色设定表+30 秒逐秒分镜的完整动画短片方案:皮克斯风 slapstick,角色一致性用 reference sheet 锁定,喜剧号可直接开工\n(来源为模型无关 prompt,模型归属为编辑指派;日期为相对时间推算)" }, { "id": "ve-031", "model": "veo-3", "category": "comedy", "title": "人体饮料机:鼻孔倒可乐的荒诞咖啡馆广告", "prompt": "First Frame Image Prompt:\n\nA hilarious, exaggerated caricature-style 3D-rendered novelty beverage dispenser modeled after the man in the reference photo. He has an oversized round head with the man's recognizable face, short black hair, thick black bushy mustache, huge wide-open surprised cartoon eyes, and an open circular \"O\" shaped mouth. He wears a shiny metallic pot-like hat on his head with a big round red button on top. His body is a massive transparent spherical bowl belly divided vertically in half - left side filled with dark cola-brown soda and ice cubes showing a Coca-Cola-style red script logo, right side filled with bright orange soda and ice cubes showing a Fanta-style orange logo. His arms are comically tiny and stubby, resting proudly on the sides of his giant belly. He wears a stretched white short-sleeved polo shirt with a small chest logo that barely fits around the round transparent torso. Two empty clear drinking glasses sit on a metal grate drip tray at the base. The dispenser stands on a warm wooden counter in a cozy rustic cafe, with a softly blurred chalkboard sign in the background reading \"Good Drink, Good Mood.\" Warm soft directional lighting, glossy glass and liquid textures, shallow depth of field, centered static frontal composition, cinematic product-shot aesthetic, highly comical and absurd tone.\n\nImage-to-Video Prompt:\n\nTheme: A comical caricature beverage dispenser modeled after a real man absurdly serves iced cola and orange soda from his mouth and nostrils in a cozy cafe\n\nVisuals: Warm rustic cafe counter, wooden surface, soft bokeh background with a chalkboard sign reading \"Good Drink, Good Mood,\" an exaggerated cartoonish man-shaped dispenser with oversized round head, bushy mustache, huge surprised eyes, open \"O\" mouth, shiny metal pot hat with a big red button, massive transparent spherical belly split into cola and orange soda halves, comically tiny stubby arms resting proudly on the belly, stretched white polo shirt, two empty clear glasses on a metal drip tray\n\nCamera: Static, locked-off centered shot with no pan, tilt, or zoom; tight framing on the dispenser and glasses throughout\n\nStyle: Hyper-realistic 3D animation mixed with caricature exaggeration, whimsical and absurd, warm color grading with rich browns, dark cola tones, vibrant orange soda, soft cinematic lighting, glossy liquid and glass textures, comedic product-shot aesthetic\n\nAction + Sound Design:\n[0-1s] A human hand enters from the top right of the frame and presses the big red button on top of the metal hat - sharp mechanical click and whirring sound\n[1-4.5s] The character's eyes widen further and his open mouth violently ejects a chaotic flurry of ice cubes, splitting into two streams that fill both empty glasses about halfway - loud clattering and spitting ice sounds\n[4.5-14s] Two solid streams of liquid begin pouring directly from the character's nostrils: dark brown cola from the left nostril into the left glass, bright orange soda from the right nostril into the right glass, steadily filling both glasses over the ice - steady pouring liquid sound with a subtle comical slurp\n[14-15s] The liquid streams stop abruptly, leaving two perfectly full iced drinks sitting on the tray while the character holds his static, proud-yet-surprised expression - soft final drip and ambient cafe quietness", "tags": [ "veo-3", "comedy", "reference", "product" ], "params": {}, "source": { "name": "sifuyik (Substack)", "url": "https://sifuyik.substack.com/p/758-viral-video-prompt", "tier": "reputable-3rd-party" }, "notes": "以真人脸做夸张 3D 漫画像饮料机的病毒创意:鼻孔倒可乐+冰块喷射,静态机位+精确到 0.5s 的音效设计,猎奇感拉满\n(来源为模型无关 prompt,模型归属为编辑指派;日期为相对时间推算)" }, { "id": "mx-011", "model": "minimax-h3", "category": "action", "title": "悬崖城市飞车追逐:一镜到底", "prompt": "Speeder chase across a cliff city (single continuous shot) From a monumental cliffside city carved into stone, the camera dives toward a tiny streak of light ripping along a narrow ledge-road. Lock-on: a speeder hugging the wall at insane speed. The camera slingshots ahead, whips back, then drops tight to the rear thrusters: heat haze, grit snapping off the ledge, warning lights flashing. A collapsing balcony rains debris; the rider snaps a last-inch swerve under a falling arch, then threads through hanging laundry lines and open windows in one fluid line. The camera darts through the same openings, staying glued to the motion. One final bend and sudden calm: the camera blasts outward into a reveal of the city opening onto a boundless waterfall-fed valley, mist turning into rainbow.\n\n@Hailuo_AI #MiniMaxH3", "tags": [ "minimax-h3", "action", "rain" ], "params": {}, "source": { "name": "X 原帖(经聚合站存档)", "url": "https://x.com/umesh_ai/status/2082499539735588916", "tier": "reputable-3rd-party" }, "notes": "一镜到底飞车追逐,镜头语言密集(slingshot/whip/drop),动作场面 prompt 范本;日期由 post ID 推算\n(日期由 X snowflake ID 推算)\n演示视频:https://raw.githubusercontent.com/callirra-ai/awesome-minimax-h3/main/videos/x-mm-h3-03.mp4" }, { "id": "mx-012", "model": "minimax-h3", "category": "documentary", "title": "街头摄影师抓拍瞬间", "prompt": "A young Western female street photographer walks through a lively downtown street and notices an elderly man sitting outside a café with his small dog. She carefully composes the candid moment through her camera, captures the photo, then turns the camera toward the viewer to proudly show the shot she just took. She smiles, says “Look at that,” then continues walking through the city. Ultra-photorealistic visuals, natural handheld documentary movement, realistic camera interaction, authentic facial expressions, accurate hand movements, realistic dog behavior, natural daylight, cinematic depth of field, continuous character consistency, immersive city ambience, premium documentary realism.", "tags": [ "minimax-h3", "documentary" ], "params": {}, "source": { "name": "X 原帖(经聚合站存档)", "url": "https://x.com/aiwithaly/status/2087102541146522089", "tier": "reputable-3rd-party" }, "notes": "纪实跟拍+相机交互+人物转镜对视,vlog 真实感写法标杆;日期由 post ID 推算\n(日期由 X snowflake ID 推算)\n演示视频:https://raw.githubusercontent.com/callirra-ai/awesome-minimax-h3/main/videos/x-mm-h3-02.mp4" }, { "id": "mx-013", "model": "minimax-h3", "category": "creative", "title": "ABC 字母启蒙动画", "prompt": "Create a 15-second animated educational video that teaches young children the letters A, B, C, and D.\n\nThe learning pattern for every letter must be:\n\nLETTER → SOUND → OBJECT → PLAYFUL ACTION → OBJECT NAME\n\nTarget audience: children ages 3 to 6.\n\nVisual style:\nUse adorable rounded 3D characters, soft pastel colors, gentle facial expressions, and simple recognizable objects. Combine this with a premium minimalist technology aesthetic featuring clean white space, elegant composition, soft studio lighting, subtle reflections, smooth gradients, rounded geometry, crisp typography, and extremely polished transitions.\n\nThe animation should feel playful and child-friendly while remaining calm, uncluttered, and beautifully designed.\n\nUse a clean off-white background with a different soft color glow behind each letter.\n\n0:00–0:01 | Introduction\nA small smiling star mascot bounces into the center of the screen.\n\nColorful letters briefly float around it.\n\nDisplay the text:\n\n“Let’s learn!”\n\nThe mascot taps the screen, creating a soft ripple that reveals the first letter.\n\n0:01–0:04 | A is for Apple\n\nShow a large uppercase “A” and smaller lowercase “a” beside it.\nUse thick, rounded, highly readable typography.\n\nThe narrator says:\n\n“A. A says ah. A is for Apple.”\n\nThe uppercase A gently inflates and transforms into a shiny red apple.\n\nIts top point becomes the apple stem, and a small green leaf unfolds from the side.\n\nThe apple gains a cute smiling face and performs one soft bounce.\n\nDisplay the word:\n\n“APPLE”\n\nHighlight the first letter A in red.\n\nAdd a soft pop and a tiny crunchy sound.\n\n0:04–0:07 | B is for Ball\n\nThe apple rolls across the screen and leaves behind a curved red trail.\n\nThe trail loops twice and forms a large uppercase “B,” with a lowercase “b” appearing beside it.\nThe narrator says:\n\n“B. B says buh. B is for Ball.”\n\nThe two rounded sections of the B expand and merge into a colorful striped ball.\n\nThe ball bounces twice with playful squash-and-stretch animation.\n\nDisplay the word:\n\n“BALL”\n\nHighlight the first letter B in blue.\n\nSynchronize each bounce with a soft musical note.\n\n0:07–0:10 | C is for Cat\n\nOn its final bounce, the ball stretches into a curved shape and becomes a large uppercase “C.”\n\nA lowercase “c” slides gently into place beside it.\n\nThe narrator says:\n“C. C says kuh. C is for Cat.”\n\nThe C rotates and becomes the curled tail of a cute orange cat.\n\nThe rest of the cat forms from soft rounded shapes.\n\nThe cat stretches, blinks, and gives one gentle wave with its paw.\n\nDisplay the word:\n\n“CAT”\n\nHighlight the first letter C in orange.\n\nAdd a quiet and friendly “meow.”\n\n0:10–0:13 | D is for Duck\n\nThe cat’s tail uncurls and transforms into the curved side of a large uppercase “D.”\n\nA lowercase “d” pops up beside it.\n\nThe narrator says:\n\n“D. D says duh. D is for Duck.”\nThe straight line of the D becomes the duck’s neck.\n\nThe curved section becomes its round yellow body.\n\nA small orange beak and two tiny wings pop into place.\n\nThe duck waddles forward, flaps its wings, and gives one cheerful quack.\n\nDisplay the word:\n\n“DUCK”\n\nHighlight the first letter D in yellow.\n\nAdd tiny water ripples beneath its feet.\n\n0:13–0:15 | Recap\n\nThe apple, ball, cat, and duck slide into four clean rounded tiles.\n\nPlace their letters above them:\n\n“A B C D”\n\nThe mascot returns and points to each object as they bounce once in sequence.\n\nNarrator:\n“A, B, C, D. Great job!”\n\nFinish with the text:\n\n“Great job!”\n\nUse a small sparkle animation and a warm musical chime.\n\nAnimation requirements:\n\nKeep each letter fully visible for a moment before it transforms.\n\nShow uppercase and lowercase versions clearly.\n\nMake every object instantly recognizable.\n\nUse smooth shape morphing so children can visually understand how the letter becomes the object.\n\nMaintain stable spelling, clean letterforms, accurate object shapes, and consistent character design.\n\nUse gentle squash-and-stretch, soft motion blur, subtle shadows, polished lighting, and precisely synchronized sound effects.\nAvoid fast camera movement, cluttered backgrounds, harsh colors, tiny text, warped letters, random symbols, duplicated objects, scary expressions, or overly complex transformations.\n\nThe final video should feel cute, educational, memorable, calming, and exceptionally polished.", "tags": [ "minimax-h3", "creative" ], "params": { "duration": "15s" }, "source": { "name": "X 原帖(经聚合站存档)", "url": "https://x.com/umesh_ai/status/2084227244533411987", "tier": "reputable-3rd-party" }, "notes": "字母→发音→物体→动作的固定学习模式+逐秒分镜,教育类动画完整模板;日期由 post ID 推算\n(日期由 X snowflake ID 推算)\n演示视频:https://raw.githubusercontent.com/callirra-ai/awesome-minimax-h3/main/videos/x-mm-h3-01.mp4" }, { "id": "mx-014", "model": "minimax-h3", "category": "comedy", "title": "办公室金正恩 NG 片段", "prompt": "(Blooper take): The Office & Kim Jong Un. Entrance. 0:00–0:04 — Medium shot at the entrance of Dunder Mifflin. MICHAEL SCOTT opens the door with a huge excited smile as KIM JONG UN steps in. Michael immediately starts singing in a playful, childish sing-song voice while pointing at him: “Kim Jong Un, Kim Jong Un… the coolest guy under the sun!” doing a little rhythmic dance with his hands. 0:04–0:07 — Michael keeps the song going and starts giving soft friendly punches to Kim’s stomach to the rhythm: “Kim Jong Un, Kim Jong Un… don’t be sad, just have some fun!” 0:07–0:10 — Kim tries hard to keep a straight face, his lips tightly pressed together, but his shoulders start shaking. He can’t hold it anymore and bursts out laughing. 0:10–0:13 — Michael freezes mid-punch, looks at Kim laughing, and immediately cracks up too. 0:13–0:15 — Both of them are laughing hard. Michael can barely stand while still holding one fist up. The video ends on the two of them completely breaking character.", "tags": [ "minimax-h3", "comedy", "dance" ], "params": {}, "source": { "name": "X 原帖(经聚合站存档)", "url": "https://x.com/techhalla/status/2084838553943449908", "tier": "reputable-3rd-party" }, "notes": "情景喜剧 blooper 结构+逐秒分镜,名人恶搞类病毒视频模板;日期由 post ID 推算\n(日期由 X snowflake ID 推算)\n演示视频:https://raw.githubusercontent.com/callirra-ai/awesome-minimax-h3/main/videos/x-mm-h3-05.mp4" }, { "id": "mx-015", "model": "minimax-h3", "category": "dialogue-driven", "title": "老友记:AI 能取代乔伊吗", "prompt": "“Joey and Chandler sit side by side on the couch with coffee cups. Joey turns to Chandler with a completely sincere, concerned expression.\n\nDIALOGUE:\n\nJOEY: “Could AI replace us?”\n\nChandler slowly looks up from his coffee. Pauses. Looks Joey dead in the eye.\n\nCHANDLER: “Joey, AI cannot replace you. Nothing can replace you.”\n\nJoey nods slowly. Visibly relieved. Processing.\n\nJOEY: “Because I’m too good?”\n\nChandler stares at him. Long beat. He looks down at his coffee, then back up.\n\nCHANDLER: “…Sure. Let’s go with that.”\n\nSTYLE: Warm 90s sitcom aesthetic, steady handheld feel, natural performances, characters in casual clothing. Joey played completely straight — zero irony. Chandler delivers every line with exhausted deadpan sarcasm. Hold on Chandler’s face after the final line for a full beat before cut.”", "tags": [ "minimax-h3", "dialogue-driven", "dialogue" ], "params": {}, "source": { "name": "X 原帖(经聚合站存档)", "url": "https://x.com/TechieBySA/status/2084600512180113820", "tier": "reputable-3rd-party" }, "notes": "双人情景对话+精确台词+表演指导(deadpan sarcasm),原生对白类 prompt 范本;日期由 post ID 推算\n(日期由 X snowflake ID 推算)\n演示视频:https://raw.githubusercontent.com/callirra-ai/awesome-minimax-h3/main/videos/x-mm-h3-08.mp4" }, { "id": "mx-016", "model": "minimax-h3", "category": "music-video", "title": "音乐卡点角色展示", "prompt": "Use @[char ref] as the strict character reference and @[audio ref] as the timing, rhythm and editing reference.\n\nKeep the character’s exact identity, proportions, hairstyle, outfit, colors and overall style consistent throughout.\n\nCreate a 15-second cinematic burst-cut video showcasing the character across 5 different environments that naturally fit their design, vibe and world.\n\nAUDIO SYNC\nSynchronize the entire edit to @[audio ref]. Cuts, camera accents, transitions and environment changes should land precisely on strong beats, half-beats and musical accents. Let audio1 control the pacing and intensity of the montage.\n\nSTRUCTURE\n- 5 environments total\n- 3 seconds per environment\n- 6 burst-cut shots per environment\n- 30 shots total\n\nEach environment must be clearly different in atmosphere, lighting, scale and visual language.\n\nShow each environment through rapid cinematic angles: wide establishing shots, aerials, low angles, side views, tracking shots, close environmental details, medium shots and hero frames.\n\nEvery cut must reveal a new angle, distance, composition or spatial relationship. Avoid repeated framing. Mix static shots, push-ins, pull-backs, tracking, orbit and crane-like movement.\n\nKeep character movement subtle and natural. The focus is environmental variety, cinematic framing and tight synchronization with audio1.\n\nHard constraints:\n- exactly 5 environments\n- exactly 6 shots per environment\n- exactly 30 shots total\n- environment changes must follow audio1’s musical phrasing\n- cuts and motion accents synchronized to audio1\n- no outfit changes\n- no character duplication\n- no morphing\n- no text or UI\n- no blurry unreadable frames\n- maintain strict character consistency", "tags": [ "minimax-h3", "music-video", "reference", "aerial", "rain" ], "params": { "duration": "15s" }, "source": { "name": "X 原帖(经聚合站存档)", "url": "https://x.com/aimikoda/status/2086553240448241802", "tier": "reputable-3rd-party" }, "notes": "音频 reference 驱动剪辑节奏(5 环境×6 镜头=30 卡点),H3 音视频统一 reference 教科书;日期由 post ID 推算\n(日期由 X snowflake ID 推算)\n演示视频:https://raw.githubusercontent.com/callirra-ai/awesome-minimax-h3/main/videos/x-mm-h3-07.mp4" }, { "id": "se-005", "model": "seedance-2.5", "category": "asmr", "title": "美食 ASMR 喜剧反转", "prompt": "Create a 30-second animated cooking sketch inside a gently rocking wooden galley. Keep one apron-wearing cook and one curious green bird recognizable in every shot. Warm lanterns illuminate tactile food surfaces.\n\n0–6s: the bird approaches a vegetable; the cook slides the chopping board away and raises an eyebrow.\n6–13s: intercut knife contact, oil ripples and bubbling sauce; synchronize each sound with the visible action.\n13–22s: the bird nudges an empty serving bowl toward the cook, then waits conspicuously.\n22–30s: the cook fills a small dish; both settle beside the window.\n\nUse expressive original character designs, quiet sea ambience and a single closing musical accent. Leave dialogue and lettering out.", "tags": [ "seedance-2.5", "asmr", "dialogue", "food" ], "params": { "duration": "30s" }, "source": { "name": "X 原帖(经聚合站存档)", "url": "https://x.com/Goodmanprotocol/status/2099186117769822462", "tier": "reputable-3rd-party" }, "notes": "9 月最新;食物音效与角色喜剧 payoff 的时间轴写法;prompt 为聚合站改编版,原作者全文见 X 原帖;日期由 post ID 推算\n(聚合站为改编版,X 原帖需登录查看;日期由 X snowflake ID 推算)\n演示视频:https://video.twimg.com/amplify_video/2099186062715404288/vid/avc1/1920x1080/H8uxwKxVsSU09_LW.mp4?tag=29" }, { "id": "se-006", "model": "seedance-2.5", "category": "social-viral", "title": "雨天自拍被小猫打断", "prompt": "Make a 30-second phone selfie beside a rain-covered window. One adult holds one tabby kitten securely. Use the supplied portrait only to preserve the person's appearance. Keep window light soft and neutral.\n\n0–7s: the person glances down at the kitten while rain continues behind them.\n7–15s: a cloth loop at the cuff attracts one paw; the person moves it aside and chuckles.\n15–23s: the kitten shifts toward the supported shoulder; the phone tilts slightly as the person adjusts their grip.\n23–30s: whiskers approach the lens, focus briefly softens, and the shot ends during a laugh.\n\nUse room sound and rain. Preserve one continuous viewpoint and the same animal.", "tags": [ "seedance-2.5", "social-viral", "rain", "portrait" ], "params": { "duration": "30s" }, "source": { "name": "X 原帖(经聚合站存档)", "url": "https://x.com/Strength04_X/status/2098256490238755226", "tier": "reputable-3rd-party" }, "notes": "一镜到底自拍+宠物互动,UGC 真实感与主体计数的写法;prompt 为聚合站改编版,原作者全文见 X 原帖;日期由 post ID 推算\n(聚合站为改编版,X 原帖需登录查看;日期由 X snowflake ID 推算)\n演示视频:https://video.twimg.com/amplify_video/2098256097547309056/vid/avc1/1920x1080/mPxvxfjQ4VcrZg5r.mp4?tag=29" }, { "id": "se-007", "model": "seedance-2.5", "category": "social-viral", "title": "双人便利店 Vlog", "prompt": "Create a 30-second evening convenience-store Vlog with two adult friends. Image 1 defines the shopper; Image 2 defines the friend filming. Keep clothing and one drink bottle consistent.\n\n0–8s: rear-camera view follows the shopper noticing a chilled drink and taking it from the refrigerator.\n8–15s: the shopper shows the bottle to the lens, then carries it to the counter; the operator speaks off-screen.\n15–23s: cut to a front-camera two-shot outside. One friend offers the other a small sip, with a visible handover.\n23–30s: they laugh and leave together, the bottle remaining with its current holder.\n\nUse modest phone shake, short conversational pauses and shop ambience. No external observer angle or music.", "tags": [ "seedance-2.5", "social-viral" ], "params": { "duration": "30s" }, "source": { "name": "X 原帖(经聚合站存档)", "url": "https://x.com/Strength04_X/status/2097968347430482177", "tier": "reputable-3rd-party" }, "notes": "双 reference 身份锁定+前后摄分工,vlog 连贯性写法;prompt 为聚合站改编版,原作者全文见 X 原帖;日期由 post ID 推算\n(聚合站为改编版,X 原帖需登录查看;日期由 X snowflake ID 推算)\n演示视频:https://video.twimg.com/amplify_video/2097967562604875776/vid/avc1/1920x1080/EB48uw1_u7MU2RJl.mp4?tag=29" }, { "id": "se-008", "model": "seedance-2.5", "category": "cinematic", "title": "善意微故事:物权连续性", "prompt": "Create a 15-second neighborhood micro-story with two fictional adults. A cyclist wearing a blue raincoat has dropped one yellow glove beside a bench. A passerby carrying a red tote notices it.\n\n0–4s: show the glove on the pavement and the cyclist looking back from the bicycle. Establish both people in one wide shot.\n4–9s: the passerby bends, picks up the glove with the right hand and approaches. Keep the red tote on the left shoulder.\n9–12s: show both hands in frame during the transfer. The cyclist closes the left hand around the glove before the passerby releases it.\n12–15s: the cyclist says 'Thanks!' and places the glove in the front basket.\n\nUse one gentle handheld move, neighborhood ambience and no music. Exactly one glove throughout; preserve its color, location and holder.", "tags": [ "seedance-2.5", "cinematic", "rain" ], "params": { "duration": "15s" }, "source": { "name": "X 原帖(经聚合站存档)", "url": "https://x.com/AIwithkhan/status/2096424933366931946", "tier": "reputable-3rd-party" }, "notes": "交接前后物权归属的显式描述,连续性控制的进阶技巧;prompt 为聚合站改编版,原作者全文见 X 原帖;日期由 post ID 推算\n(聚合站为改编版,X 原帖需登录查看;日期由 X snowflake ID 推算)\n演示视频:https://video.twimg.com/amplify_video/2096424851888300032/vid/avc1/1280x720/LYNMJekyiU57XaaC.mp4?tag=29" } ], "tier_definitions": { "official": "厂商直接拥有的域名(OpenAI/DeepMind/MiniMax/Runway/Higgsfield/Alibaba/Genmo/ByteDance 等)", "official-platform": "官方 GitHub/HuggingFace/Replicate/fal.ai 等合作部署平台", "reputable-3rd-party": "内容质量高的第三方评测/平台博客(Atlas Cloud/Atlabs/Imagine.art/CrePal 等)", "unverified-3rd-party": "SEO 农场/未经验证第三方(tryhappyhorseai/CyberLink marketing/ulazai 等)", "broken": "死链或域名失效(elser.ai)", "unknown": "无 URL 或域名未分级" } }