comfyui-mcp-server

MCP.Pizza Chef: joenorton

If you already run ComfyUI on your own machine, this puts it behind a conversation. Ask for a picture, look at the result, then ask for changes without repeating the whole description, because it remembers what it made. Songs and video are covered too, along with checking on long jobs, cancelling them, browsing earlier results, and saving a finished image into a website folder. Setup is hands-on: download the project, install its Python packages, and keep ComfyUI running alongside. No account key is needed. The last code change was February 2026.

Images/Design

Other

Use This MCP server To

Generate a picture by describing it in a sentence Ask for a tweak without retyping the whole description Check whether a long image job has finished yet Browse the images I made earlier in this session Generate a short song from lyrics and a mood Save a finished image straight into my website folder

README

ComfyUI MCP Server

Generate and refine AI images/audio/video through natural conversation

A lightweight MCP (Model Context Protocol) server that lets AI agents generate and iteratively refine images, audio, and video using a local ComfyUI instance.

You run the server, connect a client, and issue tool calls. Everything else is optional depth.

Quick Start (2–3 minutes)

This proves everything is working.

1) Clone and set up

git clone https://github.com/joenorton/comfyui-mcp-server.git
cd comfyui-mcp-server
pip install -r requirements.txt

2) Start ComfyUI

Make sure ComfyUI is installed and running locally.

cd <ComfyUI_dir>
python main.py --port 8188

3) Run the MCP server

From the repository directory:

python server.py

The server listens at:

http://127.0.0.1:9000/mcp

4) Verify it works (no AI client required)

Run the included test client:

comfyui-mcp-server FAQ

Is it still being updated?

The repository is open and not archived, but the last code change was February 2026, so it has been quiet for months.

Do I need ComfyUI already installed?

Yes. This makes nothing itself. It hands your request to a ComfyUI running on your own computer.

Does it need a key?

No. Everything runs locally, so there is no account to sign up for and nothing to pay per picture.

Can I use this to make images for a blog post?

Yes. Generate one, refine it in conversation, then have it shrunk and saved into your site's images folder.

How hard is setup?

Hands-on. Download the project, install Python packages, start ComfyUI, then start this server before connecting your app.

Which apps does it work with?

The instructions cover Cursor, Claude, and n8n, and any app that can point at a local address should work.