yesai-seedance-mcp

MCP Server for YesAI platform — Seedance video generation & Seedream image generation.

Quick Start

1. Install

No manual install needed. MCP clients run this package via npx automatically.

2. Configure

Add to your MCP config file (WorkBuddy / CodeBuddy / Cursor):

{
  "mcpServers": {
    "yesai-seedance": {
      "command": "npx",
      "args": ["-y", "yesai-seedance-mcp@latest"],
      "env": {
        "DOUBAO_API_KEY": "sk-your-api-key",
        "DOUBAO_BASE_URL": "https://your-domain.com/api/volc/v3",
        "DOUBAO_VIDEO_ENDPOINT_ID": "doubao-seedance-2-0-260128",
        "DOUBAO_IMAGE_ENDPOINT_ID": "seedream-5-0-260128"
      }
    }
  }
}

3. Use

Talk to your AI assistant in natural language:

• Text to video: "Generate a 5s 720p video: a ginger cat running through neon-lit rainy streets at night"
• Image to video: "Turn this image into a 5s video, make the character walk forward: https://example.com/photo.png"
• Text to image: "Draw a cyberpunk city skyline with neon lights at night, ultra HD"
• Query progress: "Check the status of my last video task"

Environment Variables

| Variable | Required | Description | | --- | --- | --- | | DOUBAO_API_KEY | Yes | Platform API Key | | DOUBAO_BASE_URL | Yes | Platform API base URL (must end with /api/volc/v3) | | DOUBAO_VIDEO_ENDPOINT_ID | No | Default video model ID | | DOUBAO_IMAGE_ENDPOINT_ID | No | Default image model ID |

Available Tools

| Tool | Description | | --- | --- | | generate_video | Text/image to video (async, returns task ID) | | query_video_task | Query video task status and result | | generate_image | Text/image to image (sync, returns URL) |

Supported Models

Video (Seedance): doubao-seedance-2-0-260128, dreamina-seedance-2-0-260128, doubao-seedance-2-0-fast-260128, dreamina-seedance-2-0-fast-260128, doubao-seedance-1-5-pro-251215

Image (Seedream): seedream-5-0-260128, seedream-4-0-250828, seedream-3-0-t2i-250415

Requirements

• Node.js >= 18.0.0

License

MIT