Comfy MCP Server logo

Comfy MCP Server

by MCP-Mirror

This server uses the FastMCP framework to generate images based on prompts via a remote Comfy server. It allows users to submit prompts and retrieve generated images from a ComfyUI workflow.

View on GitHub

Last updated: N/A

What is Comfy MCP Server?

This is a server that leverages the FastMCP framework to interface with a remote ComfyUI server. It enables image generation based on text prompts by utilizing a pre-defined ComfyUI workflow.

How to use Comfy MCP Server?

  1. Install Python 3.x and required packages using pip install "mcp[cli]". 2. Configure environment variables COMFY_URL, COMFY_WORKFLOW_JSON_FILE, PROMPT_NODE_ID, and OUTPUT_NODE_ID. 3. Run the comfy-mcp-server.py script. The server will then listen for requests to generate images based on provided prompts.

Key features of Comfy MCP Server

  • Remote ComfyUI interaction

  • Prompt-based image generation

  • FastMCP framework integration

  • Workflow-driven image creation

Use cases of Comfy MCP Server

  • Automated image generation pipelines

  • Integrating ComfyUI workflows into other applications

  • Generating images from text prompts programmatically

  • Creating custom image generation services

FAQ from Comfy MCP Server

What is ComfyUI?

ComfyUI is a node-based visual programming environment for creating Stable Diffusion workflows.

What is FastMCP?

FastMCP is a framework for building servers and clients that interact with machine learning models.

How do I find the PROMPT_NODE_ID and OUTPUT_NODE_ID?

These IDs can be found in the ComfyUI interface when you export your workflow as an API JSON file. Inspect the JSON to identify the correct node IDs.

What if the server fails to connect to the Comfy server?

Ensure that the COMFY_URL environment variable is correctly set and that the ComfyUI server is running and accessible.

Can I use different ComfyUI workflows with this server?

Yes, you can use any ComfyUI workflow, but you need to update the COMFY_WORKFLOW_JSON_FILE environment variable and ensure that the PROMPT_NODE_ID and OUTPUT_NODE_ID are correctly configured for the new workflow.