{ "displayName": "Volcengine", "baseUrl": "https://ark.cn-beijing.volces.com/api/v3", "apiKeyTemplate": "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx", "tokenKeyTemplate": "ark-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx-xxxxx", "models": [ { "id": "ark-code-doubao-seed-code", "name": "Doubao-Seed-Code (CodingPlan)", "model": "doubao-seed-code", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。豆包编程模型，面向Agentic编程任务进行了深度优化，具备精准的代码生成、任务调度与逻辑协同能力。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [1.2, 8, 0.24] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [2.8, 16, 0.24] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [1.4, 12, 0.24] } } ] } }, { "id": "ark-code-doubao-seed-2.0-code", "name": "Doubao-Seed-2.0-Code (CodingPlan)", "model": "doubao-seed-2.0-code", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。面向真实编程环境优化的 Coding 模型，能稳定调用 Claude Code 等常见 IDE 中的工具。模型特别优化了前端能力，在使用常见的前端框架时能有良好表现。模型支持使用 Skills，可以配合多种自定义技能使用。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [3.2, 16, 0.64] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [9.6, 48, 1.92] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [4.8, 24, 0.96] } } ] } }, { "id": "ark-code-doubao-seed-2.0-lite", "name": "Doubao-Seed-2.0-lite (CodingPlan)", "model": "doubao-seed-2.0-lite", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。兼顾生成质量与响应速度，适合作为通用生产级模型，胜任非结构化信息处理、内容创作、搜索推荐、数据分析等生产型工作。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [0.6, 3.6, 0.12] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [1.8, 10.8, 0.36] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [0.9, 5.4, 0.18] } } ] } }, { "id": "ark-code-doubao-seed-2.0-pro", "name": "Doubao-Seed-2.0-pro (CodingPlan)", "model": "doubao-seed-2.0-pro", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。旗舰级全能通用模型，适合复杂推理与长链路任务执行场景，强调多模态理解、长上下文推理、结构化生成与工具增强执行。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [3.2, 16, 0.64] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [9.6, 48, 1.92] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [4.8, 24, 0.96] } } ] } }, { "id": "ark-code-latest", "name": "Ark-Code-Latest (CodingPlan)", "model": "ark-code-latest", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。默认使用 Auto 模式，通过「效果 + 速度」双维度智能算法自动选择模型；通过开通管理页面选择或切换目标模型，切换模型后 3-5 分钟即可生效。", "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "reasoningDefault": "high", "capabilities": { "toolCalling": true, "imageInput": true } }, { "id": "ark-code-kimi-k2.7-code", "name": "Kimi-K2.7-Code (CodingPlan)", "model": "kimi-k2.7-code", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。Kimi 最新 Coding 模型，在长上下文中更可靠地遵循指令，能以更高的成功率完成编程任务，同时支持文本、图片与视频输入，思考模式，对话与 Agent 任务。", "maxInputTokens": 224000, "maxOutputTokens": 32000, "extraBody": { "thinking": { "type": "enabled" } }, "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "RMB": [6.5, 27.0, 1.3] } }, { "id": "ark-code-kimi-k2.6", "name": "Kimi-K2.6 (CodingPlan)", "model": "kimi-k2.6", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。具备强思考能力，支持多步工具调用和推理，擅长解决复杂问题，如复杂的逻辑推理、数学问题、代码编写等。", "maxInputTokens": 224000, "maxOutputTokens": 32000, "thinking": ["enabled", "disabled"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "RMB": [6.5, 27.0, 1.1] } }, { "id": "ark-code-glm-5.2", "name": "GLM-5.2 (CodingPlan)", "model": "glm-5.2", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。智谱最新旗舰模型，支持 1M 上下文窗口，长程任务效果表现突出，多项权威评测榜单成绩稳居前列。", "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "reasoningEffort": ["high", "max", "none"], "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "RMB": [8, 28, 2] } }, { "id": "ark-code-MiniMax-M3", "name": "MiniMax-M3 (CodingPlan)", "model": "MiniMax-M3", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。新一代 M 系列语言模型，在编码与智能体评测中达到行业顶尖水平，适用于 Agent 推理、工具调用、代码和长上下文任务。", "contextSize": [512000, 256000, 192000], "maxInputTokens": 448000, "maxOutputTokens": 64000, "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "RMB": [2.1, 8.4, 0.42] } }, { "id": "ark-code-MiniMax-M2.7", "name": "MiniMax-M2.7 (CodingPlan)", "model": "MiniMax-M2.7", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。M2.7 能够自行构建复杂 Agent Harness，并基于 Agent Teams、复杂 Skills、Tool 等能力，完成高度复杂的生产力任务。", "maxInputTokens": 168000, "maxOutputTokens": 32000, "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "RMB": [2.1, 8.4, 0.42] } }, { "id": "ark-code-deepseek-v4-flash", "name": "DeepSeek-V4-Flash (CodingPlan)", "model": "deepseek-v4-flash", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。DeepSeek-V4-Flash，能够提供更加快捷、经济的 API 服务。默认开启深度思考(thinking)，支持手动关闭。", "reasoningEffort": ["high", "max", "none"], "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "RMB": [1, 2, 0.2] } }, { "id": "ark-code-deepseek-v4-pro", "name": "DeepSeek-V4-Pro (CodingPlan)", "model": "deepseek-v4-pro", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/coding", "tooltip": "Coding Plan Code模型。DeepSeek-V4-Pro，Agent 能力显著增强，具备丰富的世界知识。默认开启深度思考(thinking)，支持手动关闭。", "reasoningEffort": ["high", "max", "none"], "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "RMB": [12, 24, 1] } }, { "id": "ark-plan-doubao-seed-2.0-code", "name": "Doubao-Seed-2.0-Code (AgentPlan)", "model": "doubao-seed-2.0-code", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。依托 Seed 2.0 Agent 与 VLM 能力，强化代码能力：前端出众，多语言适配，适合接入各类 AI 编程工具。默认 non-thinking，支持开启深度思考。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "thinking": ["disabled", "enabled"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [3.2, 16, 0.64] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [9.6, 48, 1.92] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [4.8, 24, 0.96] } } ] } }, { "id": "ark-plan-doubao-seed-2.0-pro", "name": "Doubao-Seed-2.0-pro (AgentPlan)", "model": "doubao-seed-2.0-pro", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。旗舰级全能通用模型，适合复杂推理与长链路任务执行场景，强调多模态理解、长上下文推理、结构化生成与工具增强执行。默认 thinking，支持关闭深度思考。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "thinking": ["enabled", "disabled"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [3.2, 16, 0.64] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [9.6, 48, 1.92] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [4.8, 24, 0.96] } } ] } }, { "id": "ark-plan-doubao-seed-2.0-lite", "name": "Doubao-Seed-2.0-lite (AgentPlan)", "model": "doubao-seed-2.0-lite", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。兼顾生成质量与响应速度，适合作为通用生产级模型，胜任非结构化信息处理、内容创作、搜索推荐、数据分析等生产型工作。默认 thinking，支持关闭深度思考。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "thinking": ["enabled", "disabled"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [0.6, 3.6, 0.12] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [1.8, 10.8, 0.36] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [0.9, 5.4, 0.18] } } ] } }, { "id": "ark-plan-doubao-seed-2.0-mini", "name": "Doubao-Seed-2.0-mini (AgentPlan)", "model": "doubao-seed-2.0-mini", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。面向低时延、高并发与成本敏感场景，提供极致的模型推理速度，适合成本和速度优先的轻量级任务。默认 thinking，支持关闭深度思考。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "thinking": ["enabled", "disabled"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [0.2, 2, 0.04] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [0.8, 8, 0.16] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [0.4, 4, 0.08] } } ] } }, { "id": "ark-plan-doubao-seed-evolving", "name": "Doubao-Seed-Evolving (AgentPlan)", "model": "doubao-seed-evolving", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。面向 Coding 与 Agent 场景，持续周级升级，以统一模型 ID 提供最新能力。支持 1M 超长上下文，具备复杂任务编排、长程规划、代码生成与工具调用能力。", "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "reasoningEffort": ["minimal", "low", "medium", "high"], "reasoningDefault": "high", "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "RMB": [6, 30, 1.2] } }, { "id": "ark-plan-code-latest", "name": "Ark-Code-Latest (AgentPlan)", "model": "ark-code-latest", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。默认使用 Auto 模式，通过「效果 + 速度」双维度智能算法自动选择模型；通过开通管理页面选择或切换目标模型，切换模型后 3-5 分钟即可生效。", "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "reasoningDefault": "high", "capabilities": { "toolCalling": true, "imageInput": true } }, { "id": "ark-plan-glm-5.2", "name": "GLM-5.2 (AgentPlan)", "model": "glm-5.2", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。智谱最新旗舰模型，支持 1M 上下文窗口，长程任务效果表现突出，多项权威评测榜单成绩稳居前列。", "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "reasoningEffort": ["high", "max", "none"], "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "RMB": [8, 28, 2] } }, { "id": "ark-plan-minimax-m3", "name": "MiniMax-M3 (AgentPlan)", "model": "minimax-m3", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。新一代 M 系列语言模型，在编码与智能体评测中达到行业顶尖水平，适用于 Agent 推理、工具调用、代码和长上下文任务。", "contextSize": [512000, 256000, 192000], "maxInputTokens": 448000, "maxOutputTokens": 64000, "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "RMB": [2.1, 8.4, 0.42] } }, { "id": "ark-plan-minimax-m2.7", "name": "MiniMax-M2.7 (AgentPlan)", "model": "minimax-m2.7", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。M2.7 能够自行构建复杂 Agent Harness，并基于 Agent Teams、复杂 Skills、Tool 等能力，完成高度复杂的生产力任务。默认 thinking，支持关闭深度思考。", "maxInputTokens": 168000, "maxOutputTokens": 32000, "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "RMB": [2.1, 8.4, 0.42] } }, { "id": "ark-plan-kimi-k3", "name": "Kimi-K3 (AgentPlan)", "model": "kimi-k3", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。Kimi 最新旗舰模型，原生支持视觉理解，并拥有 100 万 token 上下文窗口，面向软件工程、知识工作和深度推理等前沿智能场景而设计。", "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "reasoningEffort": ["max", "none"], "capabilities": { "toolCalling": true, "imageInput": true, "editTools": true }, "tokenPricing": { "RMB": [20.0, 100.0, 2.0] } }, { "id": "ark-plan-kimi-k2.7-code", "name": "Kimi-K2.7-Code (AgentPlan)", "model": "kimi-k2.7-code", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Kimi 最新 Coding 模型，在长上下文中更可靠地遵循指令，能以更高的成功率完成编程任务，同时支持文本、图片与视频输入，思考模式，对话与 Agent 任务。", "maxInputTokens": 224000, "maxOutputTokens": 32000, "extraBody": { "thinking": { "type": "enabled" } }, "capabilities": { "toolCalling": true, "imageInput": true, "editTools": true }, "tokenPricing": { "RMB": [6.5, 27.0, 1.3] } }, { "id": "ark-plan-kimi-k2.6", "name": "Kimi-K2.6 (AgentPlan)", "model": "kimi-k2.6", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。月之暗面新一代智能模型，默认 thinking，支持关闭深度思考。", "maxInputTokens": 224000, "maxOutputTokens": 32000, "thinking": ["enabled", "disabled"], "capabilities": { "toolCalling": true, "imageInput": true, "editTools": true }, "tokenPricing": { "RMB": [6.5, 27.0, 1.1] } }, { "id": "ark-plan-deepseek-v4-flash", "name": "DeepSeek-V4-Flash (AgentPlan)", "model": "deepseek-v4-flash", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。DeepSeek-V4-Flash，能够提供更加快捷、经济的 API 服务。默认开启深度思考(thinking)，支持手动关闭。", "reasoningEffort": ["high", "max", "none"], "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "RMB": [1, 2, 0.2] } }, { "id": "ark-plan-deepseek-v4-pro", "name": "DeepSeek-V4-Pro (AgentPlan)", "model": "deepseek-v4-pro", "provider": "volcengine-agent", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/plan", "tooltip": "Agent Plan 模型。DeepSeek-V4-Pro，Agent 能力显著增强，具备丰富的世界知识。默认开启深度思考(thinking)，支持手动关闭。", "reasoningEffort": ["high", "max", "none"], "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "RMB": [12, 24, 1] } }, { "id": "doubao-seed-evolving", "name": "Doubao-Seed-Evolving", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/compatible", "tooltip": "Seed-Evolving 是面向 Agent 与 Coding 场景打造的 Seed 系列模型，具备复杂任务编排、长程规划、代码生成与工具调用能力。", "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "reasoningEffort": ["minimal", "low", "medium", "high"], "reasoningDefault": "high", "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "RMB": [6, 30, 1.2] } }, { "id": "doubao-seed-2-1-turbo", "name": "Doubao-Seed-2.1-turbo-260628", "model": "doubao-seed-2-1-turbo-260628", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/compatible", "tooltip": "Doubao-Seed-2.1 是面向 Coding 和 Agent 时代打造的新一代大模型，提供 Pro 和 Turbo 两个版本，分别面向高复杂度任务探索和规模化生产场景。", "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "reasoningDefault": "high", "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "RMB": [3, 15, 0.6] } }, { "id": "doubao-seed-2-1-pro", "name": "Doubao-Seed-2.1-pro-260628", "model": "doubao-seed-2-1-pro-260628", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/compatible", "tooltip": "Doubao-Seed-2.1 是面向 Coding 和 Agent 时代打造的新一代大模型，提供 Pro 和 Turbo 两个版本，分别面向高复杂度任务探索和规模化生产场景。", "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "reasoningDefault": "high", "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "RMB": [6, 30, 1.2] } }, { "id": "doubao-seed-2-0-code", "name": "Doubao-Seed-2.0-Code-preview-260215", "model": "doubao-seed-2-0-code-preview-260215", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/compatible", "tooltip": "Doubao-Seed-2.0-Code 面向企业级编程需求优化，在 Seed 2.0 优秀的 Agent、VLM 能力基础上，特别增强了代码能力，不仅前端能力表现出众，也对企业常见的多语言编码需求做了特别优化，适合接入各种 AI 编程工具使用。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [3.2, 16, 0.64] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [9.6, 48, 1.92] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [4.8, 24, 0.96] } } ] } }, { "id": "doubao-seed-2-0-mini", "name": "Doubao-Seed-2.0-mini-260428", "model": "doubao-seed-2-0-mini-260428", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/compatible", "tooltip": "Doubao-Seed-2.0-mini-260428 新版本，能够同时输入并理解视频、图片、语音和文本四种模态，并进行跨模态的联合推理，能直接处理必须音画结合才能做出判断业务需求。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [0.2, 2, 0.04] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [0.8, 8, 0.16] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [0.4, 4, 0.08] } } ] } }, { "id": "doubao-seed-2-0-lite", "name": "Doubao-Seed-2.0-lite-260428", "model": "doubao-seed-2-0-lite-260428", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/compatible", "tooltip": "Doubao-Seed-2.0-lite-260428 新版本，支持视频、图像、音频、文本原生统一理解，是豆包大模型家族首款全模态理解模型，同时升级Agent、Coding与GUI能力。在同等算力成本下，是企业大规模、批量化部署全模态推理任务的更优性价比选择。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [0.6, 3.6, 0.12] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [1.8, 10.8, 0.36] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [0.9, 5.4, 0.18] } } ] } }, { "id": "doubao-seed-2-0-pro", "name": "Doubao-Seed-2.0-pro-260215", "model": "doubao-seed-2-0-pro-260215", "sdkMode": "anthropic", "baseUrl": "https://ark.cn-beijing.volces.com/api/compatible", "tooltip": "Doubao-Seed-2.0-pro是旗舰级全能通用模型，面向 Agent 时代的复杂推理与长链路任务执行场景。强调多模态理解、长上下文推理、结构化生成与工具增强执行。复杂指令与多约束执行能力突出，可稳定应对多步复杂规划、复杂图文推理、视频内容理解与高难度分析等场景。", "contextSize": [256000, 128000], "maxInputTokens": 224000, "maxOutputTokens": 32000, "reasoningEffort": ["minimal", "low", "medium", "high"], "capabilities": { "toolCalling": true, "imageInput": true }, "tokenPricing": { "pricing": { "RMB": [3.2, 16, 0.64] }, "tiers": [ { "contextSizeMin": 128001, "pricing": { "RMB": [9.6, 48, 1.92] } }, { "contextSizeMin": 32001, "pricing": { "RMB": [4.8, 24, 0.96] } } ] } }, { "id": "deepseek-v4-flash", "name": "DeepSeek-V4-Flash-260425", "model": "deepseek-v4-flash-260425", "tooltip": "DeepSeek-V4-Flash 是高效经济版大模型，同享百万上下文与先进架构，推理速度更快、成本更低，平衡性能与效率，适合日常问答、轻量 Agent 及高并发场景。", "reasoningEffort": ["high", "max", "none"], "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "thinkingFormat": "object", "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "RMB": [1, 2, 0.2] } }, { "id": "deepseek-v4-pro", "name": "DeepSeek-V4-Pro-260425", "model": "deepseek-v4-pro-260425", "tooltip": "DeepSeek-V4-Pro 是 DeepSeek 新一代旗舰大模型，采用 MoE 架构，原生支持百万级超长上下文，具备顶尖推理、代码与复杂 Agent 能力，适合高强度复杂任务与专业场景。", "reasoningEffort": ["high", "max", "none"], "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "thinkingFormat": "object", "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "RMB": [12, 24, 1] } }, { "id": "deepseek-v3.2", "name": "DeepSeek-V3.2-251201", "model": "deepseek-v3-2-251201", "tooltip": "模型的 EOS 时间： 2026-07-30 14:00 （UTC+8）。DeepSeek-V3.2正式版，平衡推理能力与输出长度，适合日常使用，例如问答场景和通用 Agent 任务场景", "maxInputTokens": 112000, "maxOutputTokens": 16000, "thinking": ["disabled", "enabled"], "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "pricing": { "RMB": [2, 3, 0.4] }, "tiers": [{ "contextSizeMin": 32001, "pricing": { "RMB": [4, 6, 0.4] } }] } }, { "id": "glm-4.7", "name": "GLM-4.7-251222", "model": "glm-4-7-251222", "tooltip": "模型的 EOS 时间： 2026-07-30 14:00 （UTC+8）。GLM-4.7 是智谱最新旗舰模型，更强的编程能力与更稳定的多步骤推理/执行能力。在执行复杂智能体任务提升明显，同时对话更自然，前端审美更好。", "sdkMode": "openai-responses", "maxInputTokens": 168000, "maxOutputTokens": 32000, "webSearchTool": true, "thinking": ["enabled", "disabled"], "capabilities": { "toolCalling": true, "imageInput": false }, "tokenPricing": { "pricing": { "RMB": [3, 14, 0.6] }, "tiers": [{ "contextSizeMin": 32001, "pricing": { "RMB": [4, 16, 0.8] } }] } }, { "id": "glm-5.2", "name": "GLM-5.2-260617", "model": "glm-5-2-260617", "tooltip": "GLM-5.2 是智谱面向长程自主编码场景的旗舰开源模型，全面强化了大型代码库工程、长程任务规划与工具协同能力，在多个榜单上斩获开源第一，整体表现对齐海外闭源 SOTA 模型，开发场景的成功率进一步提升。它支持真正可用的 1M 上下文，可一次性承载项目级工程，且在执行复杂智能体任务时，指令遵循与工具调用更加稳定，一次任务即可贯通需求到多端可部署产物的完整开发链路。", "sdkMode": "openai-responses", "contextSize": [1000000, 512000, 400000, 256000, 192000], "maxInputTokens": 936000, "maxOutputTokens": 64000, "webSearchTool": true, "thinking": ["enabled", "disabled"], "capabilities": { "toolCalling": true, "imageInput": false }, "extraBody": { "caching": { "type": "enabled" } }, "tokenPricing": { "RMB": [8, 28, 2] } } ] }