ComfyUI FAQs

Question 1

What is the ComfyUI MCP Server?

Accepted Answer

It's a Python server that allows AI agents to programmatically generate images using ComfyUI via the Model Context Protocol (MCP). It acts as an API to control ComfyUI workflows.

Question 2

What are the main features of this server?

Accepted Answer

It accepts image generation requests via MCP, supports dynamic parameters like prompt, width, height, and model, integrates with ComfyUI, returns image URLs, and allows workflow selection using workflow IDs.

Question 3

What are the prerequisites for using the ComfyUI MCP Server?

Accepted Answer

You'll need Python 3.10+, a locally running ComfyUI instance, and the `requests`, `websockets`, and `mcp` Python packages.

Question 4

How do I use the server to generate images?

Accepted Answer

First, run the server (`python server.py`). Then, use a client (like the provided `client.py`) to send MCP requests with your desired parameters (prompt, model, dimensions, etc.). The server will return the URL of the generated image.

Question 5

Can I use custom ComfyUI workflows with this server?

Accepted Answer

Yes! Place your API-format workflow files (exported from ComfyUI) in the `workflows/` directory and specify the workflow ID in your requests. You may need to adjust node IDs in `comfyui_client.py` for custom workflows.

ComfyUI

ComfyUI

主要功能

使用案例

主要功能

使用案例