AI Playground ComfyUI MCP
Allows interaction with AI Playground's ComfyUI backend, enabling image generation using the bundled Draft Image - Fast workflow with configurable prompts, dimensions, steps, and seeds.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@AI Playground ComfyUI MCPCreate a dreamy landscape image, 1024x1024, seed 42"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
AI Playground ComfyUI MCP
This workspace exposes AI Playground's running ComfyUI backend to VS Code and GitHub Copilot through a local Streamable HTTP MCP server.
Prerequisites
Windows 10 or 11 with PowerShell
Python 3.11, available through the Windows
pylauncherIntel AI Playground installed and able to generate images with Draft Image - Fast
The direct Python packages are mcp==2.0.0, mcp-types==2.0.0, and
httpx==0.28.1. Pip installs their transitive dependencies automatically.
Related MCP server: ComfyUI MCP Server
Setup
From the repository root, create the virtual environment expected by the launcher and install the pinned dependencies:
py -3.11 -m venv ____env
.\____env\Scripts\python.exe -m pip install --upgrade pip
.\____env\Scripts\python.exe -m pip install -r requirements.txtTo verify the installation, run the unit tests:
.\____env\Scripts\python.exe -m unittest discover -s tests -vStart
Start AI Playground and generate at least one image with Draft Image - Fast.
From this workspace, run:
.\start-server.ps1Keep that terminal open. Use the endpoint for your client:
VS Code or another Windows client:
http://127.0.0.1:8765/mcpSuperClaw with default WSL NAT:
http://host.docker.internal:8765/mcp
The server binds to
0.0.0.0so SuperClaw's WSL/Docker backend can reach it. Keep Windows Firewall enabled; this MCP server does not provide user authentication.In VS Code's MCP server view, start or refresh aipg-comfyui.
Press Ctrl+C in the server terminal to stop it.
Configuration
The server automatically uses the standard Windows AI Playground locations:
Resources:
%LOCALAPPDATA%\Programs\AI Playground\resourcesGenerated images:
%USERPROFILE%\SuperClawProjects\mediaWorkflow:
workflows\draft_image_fast.jsonin this projectMCP listener:
0.0.0.0:8765
Set any of these environment variables before starting the server to override a
default: AIPG_RESOURCES_DIR, AIPG_OUTPUT_DIR, AIPG_WORKFLOW_PATH,
AIPG_MCP_HOST, or AIPG_MCP_PORT.
For example:
$env:AIPG_OUTPUT_DIR = "D:\AI-Playground\media"
.\start-server.ps1Overrides apply to the launched server process and must refer to the matching AI Playground installation and output locations.
Tools
comfyui_status: verifies that AI Playground's ComfyUI backend is reachable.generate_image: generates a PNG with the captured Draft Image - Fast workflow.
generate_image accepts prompt, negative_prompt, width, height, steps, and seed. Width and height must be multiples of 8 between 64 and 2048. Use seed -1 for a random image. Four steps is the tuned default for the bundled LCM workflow.
Generated files remain in %USERPROFILE%\SuperClawProjects\media by
default and are also returned as MCP image content.
Troubleshooting
Start AI Playground before calling either tool.
If authentication fails after restarting AI Playground, call the tool again; the bridge discovers the newest rotating token on every invocation.
If port
8765is already in use, stop the other process before running the launcher.The ComfyUI
nodes_glsl.pyOpenGL warning in the supplied log is unrelated to this MCP bridge and does not prevent image generation.
This server cannot be deployed
Maintenance
Related MCP Connectors
Generate AI images and videos from any compatible MCP client.
Generate AI images, video, music, and sound effects, and upscale them, from any MCP client.
Use AI models for chat, image, and video generation from Claude Code and other MCP hosts.
MCP server for Flux AI image generation
Related MCP Servers
- AlicenseBqualityDmaintenanceEnables the generation of images using Together AI's models through an MCP server, supporting customizable parameters such as model selection, image dimensions, and output directory.17MIT
- AlicenseNot gradedqualityDmaintenanceEnables Claude to generate images through a local ComfyUI instance using Stable Diffusion and SDXL models via natural language. Users can trigger image generation, list available checkpoint models, and monitor the ComfyUI queue status directly from their MCP client.MIT
- AlicenseAqualityDmaintenanceMCP server that lets OpenCode (or any MCP-compatible AI coding tool) directly control ComfyUI for AI image generation.143MIT

ImageAPI AI MCP Serverofficial
AlicenseAqualityBmaintenanceMCP server for ImageAPI AI that lets AI assistants generate high-resolution images, refine prompts for free, and inspect account credits directly in development workflows and chat sessions.45 npmMIT