
ComfyUI
The ComfyUI MCP Server provides a lightweight bridge between AI agents and local ComfyUI instances, enabling generation and iterative refinement of images, audio, and video through natural conversation. It exposes tools including `generate_image`, `regenerate`, `view_image`, `get_job`, and `list_assets` that allow clients to create media with configurable parameters, poll job status, and automatically use server defaults to minimize required inputs. This solves the problem of integrating local generative AI capabilities with AI agents and applications that support the Model Context Protocol.
- GitHub stars
- 313
- tools
- 17
- last updated
- Dec 15, 2025
Tools
17 toolsPublic tool metadata for what this MCP can expose to an agent.
generate_imageGenerate images (requires `prompt`)
Generate images (requires `prompt`)
No parameter schema in public metadata yet.
generate_songGenerate audio (requires `tags` and `lyrics`)
Generate audio (requires `tags` and `lyrics`)
No parameter schema in public metadata yet.
regenerateRegenerate an existing asset with optional parameter overrides (requires `asset_id`)
Regenerate an existing asset with optional parameter overrides (requires `asset_id`)
No parameter schema in public metadata yet.
view_imageView generated images inline (images only, not audio/video)
View generated images inline (images only, not audio/video)
No parameter schema in public metadata yet.
get_queue_statusCheck ComfyUI queue state (running/pending jobs) - provides async awareness
Check ComfyUI queue state (running/pending jobs) - provides async awareness
No parameter schema in public metadata yet.
get_jobPoll job completion status by prompt_id - check if a job has finished
Poll job completion status by prompt_id - check if a job has finished
No parameter schema in public metadata yet.
list_assetsBrowse recently generated assets - enables AI memory and iteration
Browse recently generated assets - enables AI memory and iteration
No parameter schema in public metadata yet.
get_asset_metadataGet full provenance and parameters for an asset - includes workflow history
Get full provenance and parameters for an asset - includes workflow history
No parameter schema in public metadata yet.
cancel_jobCancel a queued or running job
Cancel a queued or running job
No parameter schema in public metadata yet.
list_modelsList available ComfyUI models
List available ComfyUI models
No parameter schema in public metadata yet.
get_defaultsGet current default values
Get current default values
No parameter schema in public metadata yet.
set_defaultsSet default values (with optional persistence)
Set default values (with optional persistence)
No parameter schema in public metadata yet.
list_workflowsList all available workflows
List all available workflows
No parameter schema in public metadata yet.
run_workflowRun any workflow with custom parameters
Run any workflow with custom parameters
No parameter schema in public metadata yet.
get_publish_infoShow publish status (detected project root, publish dir, ComfyUI output root, and any missing setup)
Show publish status (detected project root, publish dir, ComfyUI output root, and any missing setup)
No parameter schema in public metadata yet.
set_comfyui_output_rootSet ComfyUI output directory (recommended for Comfy Desktop / nonstandard installs; persisted across restarts)
Set ComfyUI output directory (recommended for Comfy Desktop / nonstandard installs; persisted across restarts)
No parameter schema in public metadata yet.
publish_assetPublish a generated asset into the project's web directory with deterministic compression (default 600KB)
Publish a generated asset into the project's web directory with deterministic compression (default 600KB)
No parameter schema in public metadata yet.
Sponsors
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:
# Use default prompt
python test_client.py
# Or provide your own prompt
python test_client.py -p "a beautiful sunset over mountains"
python test_client.py --prompt "a cat on a mat"
test_client.py will:
Details
- Repository
- github.com/joenorton/comfyui-mcp-server
- Category
- AI & LLM Tools