Skip to content
Stork.AIstork.ai
AI Tool

comfyui-mcp Review

ComfyUI-MCP connects AI agents to ComfyUI via the Model Context Protocol (MCP) for image, video, audio, and 3D content generation.

shipped Jul 8, 2026aifreemium
ai
comfyui-mcp — product screenshot

Why it matters

1Offers a free tier for local model usage via Ollama.
2Provides an API with documentation at /docs/llms.txt.
3Launched in public beta on June 29, 2026, connecting existing workflows with the Comfy Ecosystem.
4Supports various LLMs, including Claude, ChatGPT, Gemini, and local models via Ollama.

About comfyui-mcp

Business Model
Open Source
Platforms
macOS, Linux, Windows
Target Audience
Users of ComfyUI for AI-assisted tasks

Leadership

@artokunLinkedIn
API DocsGitHubOpen Source

Specs

API Docs

View Documentation →

GitHub

View Repository →

API Available

Yes, public API

overview

What is comfyui-mcp?

comfyui-mcp is an AI control plane tool developed by @artokun that enables AI agents and developers to generate images, video, audio, and 3D content using ComfyUI. It functions as an MCP server and autonomous sidebar agent, allowing natural language control of live ComfyUI graphs on various LLMs.

features

Key Features of comfyui-mcp

ComfyUI-MCP provides a comprehensive set of features designed to facilitate AI agent interaction with ComfyUI workflows, model management, and content generation across multiple modalities.

  • Generate images, video, and audio content.
  • Author workflows using pre-defined templates.
  • Manage AI models from platforms like HuggingFace and CivitAI.
  • Create and manage custom nodes within ComfyUI.
  • Support for local, remote, or cloud deployment configurations.
  • Natural language control of the ComfyUI graph via an autonomous sidebar agent.
  • Compatibility with various LLMs, including Claude, ChatGPT, and Gemini.
  • Integration with local LLMs through Ollama for offline operation.
  • Generate 3D models from text descriptions or image inputs.
  • Search and discover models and workflow templates.

use cases

Who Should Use comfyui-mcp?

ComfyUI-MCP targets AI agents, developers, and AI assistants seeking to automate and streamline content generation within the ComfyUI ecosystem. It simplifies complex node-based workflows through natural language interaction.

  • AI Agents: To programmatically interact with ComfyUI, generating diverse content (images, video, 3D, audio) based on prompts.
  • Developers: For building automated content pipelines, integrating ComfyUI capabilities into larger applications, and managing workflows via CLI.
  • AI Assistants: To provide users with advanced content generation capabilities without requiring direct ComfyUI canvas interaction.
  • Content Creators: For generating multiple ad variants, character designs, or script-to-shot ideation at scale through automated workflows.
  • Teams: To enable collaborative workflow authoring and reproducible generations by sharing workflow URLs.

how to use

How to Use comfyui-mcp

To utilize comfyui-mcp, users typically install the local-first control plane, connect their preferred LLM, and then issue natural language commands to drive ComfyUI workflows.

  • 1Install the comfyui-mcp project on a local macOS, Linux, or Windows system.
  • 2Configure the MCP server to connect to your ComfyUI instance.
  • 3Integrate an LLM (e.g., Claude, ChatGPT, Gemini via subscription, or local models via Ollama).
  • 4Interact with the autonomous sidebar agent using natural language prompts to generate content or modify workflows.
  • 5Utilize the comfy-cli for terminal-based workflow execution and management.
  • 6Share workflow URLs for collaborative development and reproducible content generation.

pricing

comfyui-mcp Pricing & Plans

ComfyUI-MCP operates on a freemium business model. The core comfyui-mcp project is open-source and local-first, allowing users to leverage free local LLMs via Ollama for fully offline operation. Access to cloud-based LLMs like Claude, ChatGPT, or Gemini requires an existing subscription or API key for those respective services, which are managed independently by the user.

  • Free Tier: Access to the open-source comfyui-mcp software and integration with free local models via Ollama.
  • Paid Tier: Utilizes existing user subscriptions or API keys for cloud LLMs (e.g., Claude, ChatGPT, Gemini), incurring costs directly from those third-party providers.

Pros

  • +Enables natural language control of ComfyUI graphs via AI agents.
  • +Supports a wide range of LLMs, including free local models via Ollama and cloud services like Claude, ChatGPT, and Gemini.
  • +Functions as a local-first control plane, offering privacy and offline capabilities.
  • +Facilitates automated content generation across multiple modalities: image, video, audio, and 3D.
  • +Promotes reproducibility and team collaboration through shareable workflow URLs.
  • +Simplifies complex ComfyUI workflows for users less familiar with node-based interfaces.

Cons

  • −Requires an existing ComfyUI installation and configuration.
  • −Reliance on third-party LLM subscriptions for cloud-based agent capabilities.
  • −As a public beta launched on June 29, 2026, long-term stability and comprehensive user reviews are still emerging.
  • −May have a learning curve for integrating and configuring various LLMs and ComfyUI instances.
  • −The distinction between comfyui-mcp (local project) and Comfy Cloud MCP might cause initial confusion for users.

Policies

Free Tier

Vendor website advertises a free tier.

Similar Tools

comfyui-mcp vs Competitors

ComfyUI-MCP distinguishes itself in the ecosystem of ComfyUI automation tools by offering a comprehensive, agent-native control plane rather than merely a prompt-forwarding mechanism. It provides deep interaction with the ComfyUI graph and broad LLM compatibility.

1
ComfyUI-CopilotOn StorkCompare

It is an LLM-powered plugin that uses a hierarchical multi-agent framework and extensive knowledge bases to assist users in building ComfyUI workflows with natural language.

Similar to comfyui-mcp, ComfyUI-Copilot offers natural language control and agent-based workflow generation for ComfyUI. It focuses on assisting workflow development and node suggestions, integrating directly into the UI, whereas comfyui-mcp emphasizes driving the *live* graph and managing models/custom nodes via a sidebar agent.

2
ComfyUI Workflow Generator (based on ComfyGPT research)On StorkCompare

This custom node for ComfyUI generates entire workflows from natural language descriptions using a specialized LLM and a three-stage pipeline (Generator, Validator, Builder).

This tool directly competes by enabling natural language to workflow generation within ComfyUI, similar to comfyui-mcp's agent driving the graph. It is implemented as a custom node, integrating directly into ComfyUI's node-based system, while comfyui-mcp is described as a 'control plane' with an autonomous sidebar agent.

3
ComfyUI-CrewAIOn StorkCompare

It integrates CrewAI's multi-agent collaboration framework into ComfyUI to automate and refine prompt creation and workflow design using intelligent agents.

ComfyUI-CrewAI offers multi-agent capabilities and supports various LLMs, including local models via Ollama, aligning with comfyui-mcp's agent-native and LLM flexibility. Its primary focus is on prompt creation and workflow design through agent collaboration, while comfyui-mcp acts as a broader control plane for driving the live graph.

Visit comfyui-mcp↗
Connect
𝕏
X / Twitterx.com/artokun
⌘
GitHubgithub.com/artokun/comfyui-mcp

AI Reputation Report

Is comfyui-mcp yours?

ChatGPT, Perplexity, Gemini, Claude & Grok answer buyer questions about comfyui-mcp every day. See whether they name comfyui-mcp — or send buyers to a rival.

See what AI saysfree preview