Developers · MCP

Connect Videoduck to your AI agent

The Videoduck MCP server lets Claude, Cursor, and other agents generate images, video, and audio in your Videoduck account — straight from chat. Add the server, sign in with your Videoduck account, and just ask your agent to make something. No API keys to manage.

Server endpoint

https://mcp.videoduck.ai/mcp

Add the server to your AI client

Point your client at the endpoint above — no keys or secrets to copy.

Claude Code

claude mcp add --transport http videoduck https://mcp.videoduck.ai/mcp

Cursor and other MCP clients — add it to the MCP config (Cursor: ~/.cursor/mcp.json):

Cursor / generic MCP config

{
  "mcpServers": {
    "videoduck": {
      "url": "https://mcp.videoduck.ai/mcp"
    }
  }
}

claude.ai — go to Settings → Connectors → Add custom connector and paste the endpoint above.

Sign in & approve access

The first time you connect, your client opens Videoduck in your browser. Sign in to your account and approve access — that authorizes the agent to generate on your behalf and use your credits.

claude — videoduck mcp

It's standard OAuth: nothing secret to paste or store, and you can revoke a connection from your account whenever you like. No Videoduck account yet? Create one first.

Generate

Ask your agent to create something — “generate a 9:16 product video of …”. Generation is asynchronous: it returns a job id, then the agent polls until the asset is ready and hands you the URL. Credits are spent from your Videoduck balance; failed jobs are auto-refunded.

From one prompt to a finished video

Your agent calls the Videoduck tools, polls until the render is done, and hands back the result — all from chat.

claude — videoduck mcp

Tools: generate_image · generate_video · generate_audio · get_generation_status · list_models · get_balance · upload_media

Ready? Add the connector and sign in — or sign up for Videoduck first.