Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.runcomfy.com/llms.txt

Use this file to discover all available pages before exploring further.

This guide shows how to connect the RunComfy MCP server to your AI assistant and make your first tool call.

​
Prerequisites

You need a RunComfy API token. Get yours from your Profile page.
Your API token is sent per-request in the Authorization header and is never stored by the MCP server.

​
Setup

Choose your AI assistant or MCP client below.

​
Claude Code

Run this command in your terminal: 
claude mcp add runcomfy \
  --transport streamable-http \
  https://mcp.runcomfy.com/mcp \
  --header "Authorization: Bearer <YOUR_RUNCOMFY_TOKEN>"
Replace <YOUR_RUNCOMFY_TOKEN> with your API token.

​
Cursor

Add the following to your .cursor/mcp.json file (create it if it doesn’t exist): 
{
  "mcpServers": {
    "runcomfy": {
      "url": "https://mcp.runcomfy.com/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUNCOMFY_TOKEN>"
      }
    }
  }
}

​
Windsurf

Go to Windsurf Settings > MCP and add the following configuration: 
{
  "mcpServers": {
    "runcomfy": {
      "serverUrl": "https://mcp.runcomfy.com/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUNCOMFY_TOKEN>"
      }
    }
  }
}

​
VS Code (Copilot)

Add the following to your .vscode/mcp.json file: 
{
  "servers": {
    "runcomfy": {
      "type": "http",
      "url": "https://mcp.runcomfy.com/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_RUNCOMFY_TOKEN>"
      }
    }
  }
}

​
Other MCP clients

Any client that supports Streamable HTTP transport can connect:
  • Server URL: https://mcp.runcomfy.com/mcp
  • Transport: Streamable HTTP
  • Authentication: Authorization: Bearer <YOUR_RUNCOMFY_TOKEN> header on every request

​
Verify the connection

Once configured, ask your AI assistant:
“List my RunComfy deployments”
The assistant will call list_deployments and return your deployments. If you see a list of deployment names and IDs, the connection is working.

​
Your first inference

With a deployment available, try:
“Run my [deployment name] with the prompt: a futuristic cityscape at sunset”
The assistant will:
  1. Call get_deployment to inspect the workflow’s node IDs
  2. Call submit_request with the appropriate overrides
  3. Call get_request_status to poll progress
  4. Call get_request_result to return the output URL

​
Next steps