Awesome Seedance 
Verified Seedance 2.5 / 2.0 prompt library. 463 cases checked against their original posts, 264 cross-model retests, 25 reusable templates and 60 installable AI-video Skills, drawn from goodcase.ai's 1234 verified AI cases across video, image, UI and copy. Synced daily, new cases land every day.
More verified AI cases with full prompts → GoodCase.ai
Contents
- 🚀 Start Here
- 🧩 Prompt Templates by Category
- 🧰 Skills
- 🔥 Top 30 by heat
- 🔁 Cross-model retests
- 🎬 All Prompts
- 🌐 Browse on goodcase.ai
- Why this list
- Statistics
- How to Contribute
- 🙏 Acknowledgements
- 👤 Author
🚀 Start Here
New to AI video? Follow these five steps and you will have your own clip. Nothing to install.
| Step | What to do |
|---|---|
| 1 | Pick a look in Prompt Templates by Category. Choose by the pictures. |
| 2 | Open that template (for example UGC creator review) and copy the block under Copy this. |
| 3 | Replace the [bracketed] parts with your own product, person or scene. |
| 4 | Send it to any AI chat (ChatGPT, Claude, Gemini) along with your reference images. You get back a finished prompt. |
| 5 | Paste that prompt into Seedance (Dreamina / Jimeng) and generate. If it looks off, read the template's pitfalls and re-run. |
Templates or Skills? Both are built from the same 463 verified cases. They differ in who does the work.
| Prompt templates | Skills | |
|---|---|---|
| Who it is for | Beginners, and anyone who does not want to install anything | People already working in Claude Code, Codex or another coding agent |
| How you use it | Copy, replace the [brackets], paste into any AI chat | One install command, then just tell your agent what you want |
| What it covers | 25 category templates | The same templates plus a style library, across 60 Skills and creator variants |
| What you get | One solid prompt at a time | Batches of prompts in several visual styles, picked and filled in for you |
Just browsing? Go to Top 30 by heat or all cases. Want proof that a prompt holds up? See the cross-model retests.
🧩 Prompt Templates by Category
25 templates, one per kind of clip, distilled from 448 verified cases. Pick the look you want, open its template, copy the block, swap the [bracketed] parts for your own subject, and hand it to any AI chat. Browse them all in the template index.
🧱 Structural foundations (3 templates)
Cross-cutting skeletons that almost every other template builds on: how to slice a clip into timed beats, and how to pin an identity across those beats.
| Second-by-second timeline script 4 cases filed  Split the clip into contiguous timed segments, each carrying one shot type, one main action and its own sound line. The single most load-bearing structure in the corpus. Open template · Top case | Reference image identity lock 5 cases filed  Name every reference with a stable token, enumerate what to inherit from it, and separately enumerate what must not be inherited. The inherit-nothing-else clause is what separates working locks from broken ones. Open template · Top case | Storyboard grid to video 9 cases filed  Two stages: first a single-page sheet of numbered panels from an image model, then that sheet fed to Seedance as the reference. The sheet owns order, framing and timing, and the video prompt only has to connect the panels. Open template · Top case |
📱 Realism and UGC (4 templates)
Templates that buy believability by describing camera flaws, body wear and consumer-grade optics instead of asking for quality.
| Handheld UGC vlog 92 cases filed  Buy believability with camera defects. Name a specific consumer camera era, list its flaws as requirements, and switch cinematic polish off by hand. Open template · Top case | First-person continuous take 11 cases filed  Bodycam, GoPro, FPV and handlebar POV. The camera is mounted on a body, so its motion has to be derived from that body, and every cut has to be declared by hand. Open template · Top case | Early-2000s DV home video 17 cases filed  The period feel comes from camera defects and one tiny everyday incident. Write the camcorder's autofocus hunting, exposure shifts and handheld shake as an explicit list, keep the story small, and the clip reads as a real old tape. Open template · Top case |
| Pets and animals as the lead 13 cases filed  The animal is the lead and a phone is the only camera. Lock the count to exactly one, keep the animal behaving like an animal, and let the payoff come from it closing in on the lens. Open template · Top case |
🛒 Commercial and product (4 templates)
Ad-shaped structures where a product has to survive macro shots, hand contact and a hero frame without deforming.
| UGC creator review with spoken lines 11 cases filed  A creator unboxes, handles and endorses a product on camera. Two independent locks are needed — one on the person, one on the product — and the spoken lines are welded into the actions. Open template · Top case | Cinematic product commercial shot list 27 cases filed  A polished 8 to 20 second ad: a stated commercial aesthetic up front, a numbered or timed shot breakdown in the middle, a hero frame at the end, and a keyword tail. Open template · Top case | Process and transformation montage 16 cases filed  Cooking steps, renovation timelapse, blueprint-to-house, miniature city assembly. The craft here is declaring what must not change, then ordering the change spatially. Open template · Top case |
| Fashion lookbook and portrait film 13 cases filed  One person, one look, a handful of places. A head-to-toe appearance lock carries the whole clip, and each scene gets one location, one gesture and one kind of light. Open template · Top case |
🎭 Narrative and performance (5 templates)
Templates where the payload is a story beat or a line of dialogue rather than a look.
| Dialogue and performance beats 8 cases filed  Declare the spoken language, tag the speaker, write the reaction as a causal chain rather than a list of expressions, and close each beat with an explicit end state. Open template · Top case | Cinematic narrative short 30 cases filed  Multi-act storytelling in 15 to 60 seconds. Titled acts, a character card ahead of the acts, and a reveal written as a concrete image rather than as a promise of surprise. Open template · Top case | Cinematic travel vlog montage 12 cases filed  One traveller moves through a place scene by scene, each scene carrying its own timecode, its own location and one short line. The polish is bought with film grain and golden-hour light. Open template · Top case |
| Twist-ending comedy skit 22 cases filed  A short skit where every beat is laid down to serve one punchline. The joke has to land on something visible, and the absurdity only works when the camera and the physics stay dead serious. Open template · Top case | Horror and suspense 11 cases filed  Every shot carries its own timecode and shows one visible change on a body. The dread comes from the chain — a look, veins, a bite, the next person — and the ending seals a door without settling anything. Open template · Top case |
🎨 Stylized animation (3 templates)
Non-photoreal looks that collapse into generic CG unless the style is specified as measurable parameters plus an exclusion list.
| Anime and stylized style lock 25 cases filed  Specify the drawing style as measurable parameters, then attach an exclusion list of the neighbouring styles it must not fall into. Without the exclusion list, anime collapses into a generic 3D face. Open template · Top case | Stop motion and stepped cadence 9 cases filed  Stop motion is a timing spec before it is a look. Pin the frame rate and the hold count, name the craft material, and ban the three things that silently smooth it away. Open template · Top case | 3D cartoon character short 15 cases filed  One small anthropomorphic character carries the whole film. Spell out its looks part by part, write the style as render settings you can measure, and cut the timeline so each block holds a single action goal. Open template · Top case |
💥 Action, dance and effects (6 templates)
Templates driven by body mechanics, beat placement or a physics set-piece rather than by scene description.
| Combat choreography 40 cases filed  Fights read as real when the prompt specifies biomechanics, contact points and an attack chain. Adjectives like epic produce two people swinging at air. Open template · Top case | Beat-synced music video 12 cases filed  Derive beat anchors from BPM, pin every cut, hair flip and formation change to a real downbeat, and constrain the backup dancers so they never steal the visual centre. Open template · Top case | Time freeze and rewind set piece 5 cases filed  A five-beat skeleton — normal, collision, freeze at the peak, orbit, precise rewind — with exactly one character exempt from the freeze. The corpus contains the same author reusing this skeleton with a different physical material, which is direct evidence that it transfers. Open template · Top case |
| Cars and vehicles at speed 10 cases filed  The machine has to stay one machine while the camera does all the work. Lock the vehicle part by part, then fill the runtime with a numbered cut list that moves the camera every second. Open template · Top case | Epic fantasy and sci-fi spectacle 21 cases filed  Monsters, dragons, world reveals. Each entity gets its own definition block, the shots are cut by timecode, and scale is bought with reference objects and low angles rather than words like massive. Open template · Top case | Sports and extreme stunts 10 cases filed  The clip lives or dies on the action loop. Write every link from run-up to landing in order, name the physics you want by part, and spend the negative list on flying, hovering and teleporting. Open template · Top case |
🧰 Skills
A Skill is an installable instruction pack for coding agents (Claude Code, Codex and friends). Install one and just tell your agent what you want: it picks the template, fills the structure, and pulls from the style library to give you several looks at once. 29 Skills below, plus 31 creator variants that carry one creator's signature style.
This repo's Skill. Every category template on this page plus the style library, so your agent writes structured Seedance prompts inside your editor. npx seedance-prompt-library install |
Every beat serves one punchline that lands on something visible; absurd setups shot with a straight-faced camera. npx skills add LearnPrompt/awesome-seedance --skill seedance-meme-comedy |
Dragons, titans, world reveals: one definition block per entity, timecoded cuts, scale bought with low angles and reference objects. npx skills add LearnPrompt/awesome-seedance --skill seedance-epic-fantasy-scifi | ||||||||||||
Period feel bought with a camcorder-flaw list and one small everyday event; the clip plays like a real old tape. npx skills add LearnPrompt/awesome-seedance --skill seedance-retro-dv-home-video |
One appealing character carries the clip: parts-list design lock, measurable render look, one action goal per timeline block. npx skills add LearnPrompt/awesome-seedance --skill seedance-3d-cartoon |
Two stages: first a single-page sheet of numbered panels from an image model, then that sheet fed to Seedance as the reference. The sheet owns order, framing and timing, and the video prompt only has to connect the panels. npx skills add LearnPrompt/awesome-seedance --skill seedance-storyboard-grid-to-video | ||||||||||||
One traveller moves through a place scene by scene, each scene carrying its own timecode, its own location and one short line. The polish is bought with film grain and golden-hour light. npx skills add LearnPrompt/awesome-seedance --skill seedance-travel-city-walk |
The animal is the lead and a phone is the only camera. Lock the count to exactly one, keep the animal behaving like an animal, and let the payoff come from it closing in on the lens. npx skills add LearnPrompt/awesome-seedance --skill seedance-pet-animal |
The machine has to stay one machine while the camera does all the work. Lock the vehicle part by part, then fill the runtime with a numbered cut list that moves the camera every second. npx skills add LearnPrompt/awesome-seedance --skill seedance-car-vehicle | ||||||||||||
One person, one look, a handful of places. A head-to-toe appearance lock carries the whole clip, and each scene gets one location, one gesture and one kind of light. npx skills add LearnPrompt/awesome-seedance --skill seedance-fashion-lookbook |
Every shot carries its own timecode and shows one visible change on a body. The dread comes from the chain — a look, veins, a bite, the next person — and the ending seals a door without settling anything. npx skills add LearnPrompt/awesome-seedance --skill seedance-horror-suspense |
The clip lives or dies on the action loop. Write every link from run-up to landing in order, name the physics you want by part, and spend the negative list on flying, hovering and teleporting. npx skills add LearnPrompt/awesome-seedance --skill seedance-sports-extreme | ||||||||||||
Lets your agent search the whole goodcase.ai library live: every model, full prompts, heat, stability scores and retest baselines. npx skills add LearnPrompt/goodcase-lite --skill goodcase |
First-person framing, handheld motion and everyday detail that make a clip feel like someone was really there. npx skills add LearnPrompt/goodcase-lite --skill pov-vlog-presence10 creator variants |
Starts from the selling point and lays out the opening hook, usage scenes and the closing brand shot. npx skills add LearnPrompt/goodcase-lite --skill product-ad-shot-design5 creator variants | ||||||||||||
Locks character design, movement language and material look across multi-shot animation. npx skills add LearnPrompt/goodcase-lite --skill animation-style-consistency2 creator variants |
Writes characters, direction of motion, rhythm and shot hand-offs as an action sequence you can reproduce. npx skills add LearnPrompt/goodcase-lite --skill action-continuity-choreography1 creator variant |
Tells a making-of, a makeover or a before-and-after as a sequence of clear, satisfying stages. npx skills add LearnPrompt/goodcase-lite --skill process-transformation-story | ||||||||||||
Period feel bought with a camcorder-flaw list and one small everyday event, so the clip plays like a real old tape. npx skills add LearnPrompt/goodcase-lite --skill retro-dv-home-video2 creator variants |
Dragons, titans, world reveals: one definition block per entity, timecoded cuts, and scale bought with low angles and reference objects. npx skills add LearnPrompt/goodcase-lite --skill epic-fantasy-scifi-spectacle4 creator variants |
Every beat serves one punchline that lands on something visible; absurd setups shot with a straight-faced camera. npx skills add LearnPrompt/goodcase-lite --skill twist-comedy-skit1 creator variant | ||||||||||||
One appealing character carries the clip: a parts-list design lock, a measurable render look, and one action goal per timeline block. npx skills add LearnPrompt/goodcase-lite --skill 3d-cartoon-character-short1 creator variant |
One traveller moves through a place scene by scene, each with its own timecode, location and one short line; polish comes from film grain and golden-hour light. npx skills add LearnPrompt/goodcase-lite --skill travel-city-walk2 creator variants |
The animal is the lead and a phone is the only camera: lock the count to one, keep it behaving like an animal, let the payoff be it closing in on the lens. npx skills add LearnPrompt/goodcase-lite --skill pets-and-animals-lead1 creator variant | ||||||||||||
One machine from first frame to last while the camera does the work: lock the vehicle part by part, then fill the runtime with a numbered cut list that moves every second. npx skills add LearnPrompt/goodcase-lite --skill vehicles-at-speed |
One person, one look, a handful of places: a head-to-toe appearance lock carries the clip, each scene gets one location, one gesture and one light. npx skills add LearnPrompt/goodcase-lite --skill fashion-lookbook-portrait1 creator variant |
Every shot carries its own timecode and one visible change on a body; dread comes from the chain, and the ending seals a door without settling anything. npx skills add LearnPrompt/goodcase-lite --skill horror-suspense-short1 creator variant | ||||||||||||
It all rides on the action loop: write every link from run-up to landing in order, name the physics you want by part, and spend the negative list on flying, hovering and teleporting. npx skills add LearnPrompt/goodcase-lite --skill sports-extreme-stunts |
Two stages: first a single sheet of numbered panels from an image model, then that sheet fed to Seedance as the reference; order, framing and timing live on the sheet. npx skills add LearnPrompt/goodcase-lite --skill storyboard-grid-to-video |
Every install line works with the skills CLI. Skills named seedance-… live in this repo under agents/skills: the library Skill carries every template, the single-kind Skills carry one template each with their own case evidence, all regenerated daily from the same data. npx seedance-prompt-library install also drops the library Skill straight into Claude Code and Codex. More Skills across image, coding and writing live on goodcase.ai.
🔥 Top 30 by heat
The 30 hottest cases across all versions. prompt opens the full entry in the gallery, source opens the creator's original post.
🔁 Cross-model retests
As far as we know, this is the first public prompt library that re-runs its video prompts on a second model at scale and publishes the result either way. 254 of the cases here have been re-run (264 runs so far), each with a verdict, a judge score and the generated output. A prompt that only ever worked once, for its author, on one model, is a screenshot; a prompt that survives a re-run is a method.
| Model | Runs | Reproduction rate |
|---|---|---|
| MiniMax H3 Max 768p | 253 | 73% |
| MiniMax H3 768p | 11 | 82% |
Verdicts across all runs: ✅ 193 reproduced · ⚠️ 68 degraded · ❌ 3 failed. Runs without a final score show as score n/a. Per-case verdicts, scores and output videos are on each case's goodcase.ai page; the model labels and batch dates are explained under Statistics.
Same prompt, second model. Three examples, including one that did not hold up:
| Case | Original (Seedance) | Retest (second model) | Verdict |
|---|---|---|---|
| Seoul Summer Night Vlog 2.5 · heat 99 |  |  MiniMax H3 Max 768p ▶ output video | ✅ reproduced (score 75.7) |
| Mini DV Coffee ASMR Vlog 2.5 · heat 99 | |  MiniMax H3 Max 768p ▶ output video | ✅ reproduced (score 77.3) |
| Frozen Time and Rewind in a 1950s Diner 2.5 · heat 100 | |  MiniMax H3 Max 768p ▶ output video | ⚠️ degraded (score 63.9) |
Every run costs real inference money: over US$300 across 264 runs so far, at list price with no discounts, which is what anyone else would pay to reproduce them. We publish the result whether or not it flatters the prompt. Want Kling, Veo, Hailuo or the Seedance 2.5 API added to the retest matrix? Sponsor a batch → or write to carl@goodcase.ai.
🎬 All Prompts
All 463 cases, with full prompts, live in the gallery under docs/ (sharded so GitHub renders every page). Start from the gallery index, or jump to a version:
- Seedance 2.5 - 258 cases: Part 1 (cases 1–85) · Part 2 (cases 86–176) · Part 3 (cases 177–246) · Part 4 (cases 247–258).
- Seedance 2.0 - 205 cases: Part 1 (cases 1–101) · Part 2 (cases 102–186) · Part 3 (cases 187–205).
🌐 Browse on goodcase.ai
This README is an index. The full experience lives at goodcase.ai: search across every case, the heat leaderboard, the stability ranking, per-case retest logs with output videos, and the installable Skills that grow out of the cases. Every entry here links back to its goodcase.ai record.
Why this list
Human-verified against the source. Every prompt here was checked against the creator's original post. Prompts reverse-engineered from the output video only, with no source and no submission, are rejected outright, per goodcase.ai's collection standards (in force since 2026-08-05).
Re-run on a second model. Most cases have been re-generated on another video model, with the verdict, score and output published. See Cross-model retests.
Full provenance on every entry. Author, original post link, publish date, and a heat score, a relative percentile among published cases on the same platform. If it didn't rank, it isn't here.
Ships as an installable Agent Skill. npx seedance-prompt-library install drops a template library straight into Claude Code / Codex so your agent writes Seedance prompts from proven structures, not guesses.
Statistics
| Metric | Value |
|---|---|
| Seedance cases in this repo | 463 |
| Seedance 2.5 | 258 |
| Seedance 2.0 | 205 |
| Unique authors | 166 |
| Re-run on other models | 254 cases / 264 runs |
| Stability score (measured) | 252 cases / avg 77.9 |
| Last updated | 2026-09-19 |
| goodcase.ai, all categories | 1234 cases / 352 creators |
| goodcase.ai, AI video | 625 cases |
Retest batch note: Retest runs carry two MiniMax labels on purpose. The 2026-08 batches (runIds video-minimax-h3-768p-20260809-phase1 / -20260811-top-heat) ran on Flova's MiniMax H3 768p; the 2026-09 batches (video-minimax-h3-768p-20260906-phase1 / -20260907-phase1) ran on fal.ai's minimax/h3-max/text-to-video endpoint, which only existed from 2026-09-03. They are different model tiers on different platforms, so the labels are kept distinct rather than merged. The 2026-08 entries show finalScore: null because a later human-review pass overwrote their structured evidence notes with a prose verdict; those rows carry a human-reviewed verdict instead, and the original judge scores remain in the private eval evidence manifest.
Each case is counted once; a case tagged with several Seedance versions counts under the highest one.
How to Contribute
New prompt cases go through goodcase.ai's review pipeline so provenance and heat score stay verifiable: submit at goodcase.ai/submit (collection standards: goodcase.ai/standards). Prefer GitHub? Open a pull request that adds one JSON file under submissions/ following submissions/TEMPLATE.json; a maintainer pushes it through the same review, and it lands in data/ on the next export.
Pull requests are welcome for template fixes and new templates in data/templates-local.json, generator and Skill code under scripts/ and agents/, and corrections to English titles or summaries. README.md, README_zh.md, docs/ and the Skill reference are generated from data/, so please don't hand-edit them: change the source, run npm test && npm run generate, and commit the regenerated files in the same PR. See contributing.md for the submission standard, what gets rejected, and how the generator works. This project follows the code of conduct.
🙏 Acknowledgements
This project's format and Skill-packaging approach were shaped by:
- freestylefly/awesome-gpt-image-2 - Template-library + installable-Skill + marketplace pattern.
- YouMind-OpenLab - README-as-gallery with per-entry attribution.
- goodcase.ai - The source of every case, heat score and retest in this repository.
👤 Author
Carl (@卡尔的AI沃茨) started GoodCase and maintains this list. Six years of large-model R&D at ByteDance, Alibaba and Baidu before turning to AI video, then a run of AI shorts that made it onto Chinese TV: I Am the Tile Cat (CCTV), When the Stars Shine: AI Bonus (Hunan TV), I Miss You So Much (Zhejiang TV) and Does Birth Decide Fate? I Refuse (Bilibili), each past a million plays. 500k+ followers across platforms, speaker at WAIC and GDPS, guest lecturer on AI at the Communication University of China.
Every case here is something he or the GoodCase reviewers checked against the original post. Follow along on X @aiwarts or the WeChat official account 卡尔的AI沃茨 (Chinese). Questions, corrections and collaborations: carl@goodcase.ai.
 Scan to join the GoodCase Feishu group (Chinese): new cases, retest results, template drops. |
Copyright & Takedown Notice
This repository carries three kinds of material under three different terms.
Code (generator scripts, Agent Skill, tooling) is released under the MIT License, see the LICENSE file. The MIT license covers the code only.
Curation (selection, organization, statistics, template extraction, summaries written by us) is licensed under CC BY 4.0. Reuse it with attribution to awesome-seedance / goodcase.ai.
Prompts and media remain the copyright of their original creators. Prompt text, creator summaries, poster images and video references are quoted from publicly published posts for documentation and study; every entry links back to its original source and to its goodcase.ai record. Nothing here grants a license to the underlying prompt or media beyond what the original post allows.
Takedown process. If you are a rights holder and want an entry removed or corrected, open a GitHub issue with the entry's slug (from its goodcase.ai URL) and the original source link, or contact goodcase.ai directly. Requests are verified against the original post and honored on verification; the entry is removed from data/ and disappears from every generated file on the next regeneration.
Star History
License & Reuse
Code in this repository is open source under the MIT License: use it, modify it, build on it, keep the license notice. Curation is CC BY 4.0; prompts and media stay with their creators. Details above under Copyright & Takedown Notice.