MeiGen AI Design MCP Server

Content

MeiGen AI Design MCP Server is a Node-based MCP server and CLI for creative image and video workflows. It exposes tools for searching an inspiration gallery, retrieving prompt examples, enhancing prompts, listing models, generating images, generating videos, managing ComfyUI workflows, and storing local preferences.

Use it when Claude needs a dedicated design-generation interface rather than a generic image prompt. MeiGen can route generation through MeiGen Cloud, OpenAI-compatible image APIs, or local ComfyUI workflows, so provider setup and cost review matter before calling generation tools.

Source Review

• https://github.com/jau123/MeiGen-AI-Design-MCP
• https://raw.githubusercontent.com/jau123/MeiGen-AI-Design-MCP/main/README.md
• https://registry.npmjs.org/meigen
• https://raw.githubusercontent.com/jau123/MeiGen-AI-Design-MCP/main/LICENSE
• https://raw.githubusercontent.com/jau123/MeiGen-AI-Design-MCP/main/package.json
• https://raw.githubusercontent.com/jau123/MeiGen-AI-Design-MCP/main/src/server.ts
• https://raw.githubusercontent.com/jau123/MeiGen-AI-Design-MCP/main/src/config.ts
• https://raw.githubusercontent.com/jau123/MeiGen-AI-Design-MCP/main/src/tools/generate-image.ts
• https://raw.githubusercontent.com/jau123/MeiGen-AI-Design-MCP/main/src/tools/generate-video.ts
• https://raw.githubusercontent.com/jau123/MeiGen-AI-Design-MCP/main/src/lib/upload.ts

These sources were reviewed on 2026-06-06. Prefer the live repository, README, npm metadata, license, package metadata, server registration, config loader, image/video generation tools, and upload helper for current behavior.

Features

• Run as an npm stdio MCP server with npx -y meigen.
• Search a curated prompt and inspiration gallery.
• Retrieve prompt examples and associated image metadata.
• Enhance short creative ideas into richer image prompts.
• List available image and video models across configured providers.
• Generate images through MeiGen Cloud, OpenAI-compatible APIs, or local ComfyUI workflows.
• Generate videos through the MeiGen provider with model, duration, resolution, aspect-ratio, first-frame, last-frame, and reference-video options.
• Compress and upload local reference images for API providers.
• Manage user preferences, preferred styles, favorite prompts, and recent generations.
• Import, view, modify, delete, and use ComfyUI workflow templates.
• Use the same package as a standalone CLI for scripted generation.

Installation

Add the server to an MCP client with npm:

{
  "mcpServers": {
    "meigen": {
      "command": "npx",
      "args": ["-y", "meigen"],
      "env": {
        "MEIGEN_API_TOKEN": "{meigen-api-token}"
      }
    }
  }
}

The inspiration, prompt-enhancement, model-listing, workflow, and preference tools can be useful before generation credentials are configured. For image or video generation, configure one of the provider paths supported by upstream: MeiGen API token, OpenAI-compatible API credentials, or local ComfyUI workflows.

Use Cases

• Search a prompt gallery for visual direction before drafting a campaign image.
• Expand a terse product, logo, poster, or social-media idea into a stronger prompt.
• Generate a small set of reviewed image directions through a configured provider.
• Generate a short video only after confirming model, duration, resolution, and expected cost.
• Use local ComfyUI workflows from Claude while keeping workflow files under review.
• Save preferred visual style, aspect ratio, model, and favorite prompts for repeated design sessions.
• Use the package as a CLI in controlled scripts after reviewing secrets and output paths.

Safety and Privacy

Treat MeiGen as a generation and upload bridge. Local reference images may be compressed and uploaded to a public URL before API-provider generation, while ComfyUI workflows can run locally and may depend on local models, custom nodes, or GPU resources. Review source images, prompts, workflow JSON, and model choices before approving generation.

Generated image and video calls can spend credits or hit paid APIs. The server source instructs clients not to parallelize videos and warns that timed-out video jobs may still be running. Confirm expensive, slow, multi-image, video, high-resolution, or reference-video calls before use, and review rights, likeness, brand, platform, and client confidentiality obligations before sharing outputs.

Duplicate Check

No jau123/MeiGen-AI-Design-MCP, MeiGen AI Design MCP Server, meigen package entry, or matching source URL entry was found in content/mcp or README.md.

• Source repositorygithub.com 2026-07-21T07:17:01+00:00
• Documentationraw.githubusercontent.com
• Submitted by oktofeesh12026-06-06