M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
M

MeiGen AI Design MCP

Community
Open-source image and video generation MCP server for major AI coding tools
How to install ↓ Compare GitHub source repository ↗
Category
Other #106 of 230
Stars
★ 1.8k Very popular
Transport
stdio (local process)
Runtime
Node.js · Hosted (no local runtime)
Credentials
Optional API key
License
MIT
Last commit
Tools
8
55FMRS · C

Based on the supplied repository material, this is an MIT-licensed open-source stdio MCP server for AI image and video generation. The evidence documents 8 tools, three backend types, 1,446 prompt templates, and integrations with multiple AI clients. It is not maintained by the upstream model or MCP protocol owner, so official is false.

Strongest · Documentation 14/20 Weakest · Reliability 8/20

Reliability
8/20
Security and permissions
11/20
Maintenance
9/20
Documentation
14/20
Setup experience
13/20
Why each score
Reliability 8/20
Only the README is available; there is no verifiable server manifest, source code, test suite, or CI execution evidence. Therefore MCP initialization, tool-list handshake, and consistency between declared and actual behavior cannot be confirmed. The README describes eight tools, multiple backends, and long-running video jobs, but provides little evidence for error handling, dependency pinning, timeouts, retries, or edge cases. Static calibration caps the score at 12, so points are deducted.
Security and permissions 11/20
The README discloses API-key configuration, external APIs, Cloudflare R2 uploads, public URLs, 24-hour retention, and the local ComfyUI boundary, and states that there is no telemetry. Points are deducted because there is no source-level evidence for permission scope, secret protection, upload validation, rate limiting, dangerous-operation confirmation, or endpoint trust verification. No stated security red line is evident.
Maintenance 9/20
The project states an MIT license and is not archived, and it shows strong community-use signals plus five open issues. Stars are discovery signals and add no points. There is no verifiable commit history, release record, CI, dependency-update policy, maintainer response commitment, or security-reporting channel, so the score remains moderate-to-low.
Documentation 14/20
The README covers installation for multiple clients, configuration files, environment variables, authentication, provider selection, tool inventory, output paths, privacy, and some limitations. It is comparatively thorough. Points are deducted because the manifest is unavailable, tool parameters and error semantics are incomplete, pricing is delegated to an external page, and there is no verifiable troubleshooting guide, compatibility matrix, or test evidence.
Setup experience 13/20
The README provides concrete steps and configuration examples for Claude Code, Cursor, VS Code, Windsurf, Roo Code, OpenClaw, Hermes, and generic MCP clients, making the intended path fairly clear. Points are deducted for reliance on npx, client plugins, external services, credentials, version assumptions, and special video timeouts. Without source, tests, or an actual connection verification, the score cannot approach full marks.

Static review · not runListed 2026-07-29

Read the FMRS scoring method →

Fit and risk

What it can accessWrites / deletes local filesUses the network

Best for

  • Users creating images or videos inside AI coding tools
  • Design workflows that benefit from curated prompts and prompt enhancement
  • Users with a local GPU or an image-generation API backend

Not for

  • Users who only need text or conventional coding capabilities
  • Users who do not want prompts or reference images sent to third-party cloud APIs unless using local ComfyUI
  • Users without a configured generation backend or API credentials who need image or video generation

Required permissions

  • Run npx and access the npm package
  • Provide an API key or token for MeiGen Cloud or an OpenAI-compatible API
  • Access the configured ComfyUI service when using local generation
  • Send prompts and, where applicable, reference images to the configured service; local reference images may be uploaded
  • Write configuration and generated results to local configuration, image, and video directories

Risks and side effects

  • Cloud backends receive prompts and reference images, with privacy handling determined by the provider
  • Reference images are uploaded to Cloudflare R2 and expire automatically after 24 hours
  • Image and video generation may incur service costs or be limited by model, resolution, and quality capabilities
  • Local files may be compressed and uploaded when using non-ComfyUI backends
  • Video generation can take several minutes and may fail if client timeouts are too short

Setup

Before you start

Runtime：Node.js · Hosted (no local runtime)

MEIGEN_API_TOKEN optionalsecret MeiGen platform token, created at meigen.ai → Settings → API Keys (starts with meigen_sk_); required for image/video generation.
OPENAI_API_KEY optionalsecret API key from any OpenAI-compatible provider (e.g. Together AI) for bring-your-own-API mode.
Other optional settings (8)
OPENAI_BASE_URL optional Base URL of the OpenAI-compatible API, e.g. https://api.together.xyz/v1.
OPENAI_MODEL optional Model ID supported by your endpoint, e.g. black-forest-labs/FLUX.1-schnell.
COMFYUI_URL optional Local ComfyUI server URL, default http://localhost:8188.
MEIGEN_OUTPUT_DIR optional Overrides the local save directory for generated images, default ~/Pictures/meigen.
MEIGEN_VIDEO_OUTPUT_DIR optional Overrides the local save directory for generated videos, default ~/Movies/meigen.
XDG_PICTURES_DIR optional Linux-only default image save location used when MEIGEN_OUTPUT_DIR is unset.
XDG_VIDEOS_DIR optional Linux-only default video save location used when MEIGEN_VIDEO_OUTPUT_DIR is unset.
UPLOAD_GATEWAY_URL optional Custom S3/R2 presign upload gateway endpoint for reference image uploads.

Run the meigen package with npx and add the example configuration to a compatible MCP client configuration file. Before generating images or videos, configure local ComfyUI, MeiGen Cloud, or an OpenAI-compatible API with /meigen:setup or the documented environment variables. The first npx run requires network access.

.mcp.json 
{"mcpServers":{"meigen":{"command":"npx","args":["-y","[email protected]"],"env":{"MEIGEN_API_TOKEN":"meigen_sk_..."}}}}

Shown for Claude Code. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json
{
  "servers": {
    "meigen": {
      "command": "npx",
      "args": [
        "-y",
        "[email protected]"
      ],
      "env": {
        "MEIGEN_API_TOKEN": "meigen_sk_..."
      }
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal
claude mcp add meigen -e MEIGEN_API_TOKEN=meigen_sk_... -- npx -y [email protected]

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

The client's tool list should show search_gallery, generate_image, list_models, etc.; try "Search for some creative inspiration" — this works without any API key and proves the connection.

Troubleshooting

  1. Restart Claude Code or the client after installation or configuration changes
  2. Run /meigen:setup to check provider, token, API URL, and workflow settings
  3. Check MEIGEN_API_TOKEN, OPENAI_API_KEY, OPENAI_BASE_URL, OPENAI_MODEL, and COMFYUI_URL; environment variables take priority
  4. For first-run npx downloads or video generation, check connect_timeout and timeout settings
  5. Use list_models to inspect supported models, resolutions, and quality tiers
  6. With ComfyUI, verify that the service is running at the configured URL and that the workflow uses API format

Things to try

Once connected, you can ask your AI assistant things like:

  • Search for some creative inspiration
  • Enhance this brief idea into a professional image prompt: a calico cat in a sunlit kitchen
  • Generate a 1:1 tech logo with midjourney-v8.1
  • List all available image and video models

Tools 8

search_gallery read-only
Search 1,446 curated trending prompts with visual previews.
get_inspiration read-only
Retrieve the full prompt, all images, and metadata for a gallery entry.
enhance_prompt read-only
Transform a brief idea into a professional image prompt.
list_models read-only
List models available from configured providers.
comfyui_workflow writes
List, view, import, modify, or delete ComfyUI workflow templates.
manage_preferences writes
Remember preferred styles, aspect ratios, models, and favorite prompts.
generate_image writes
Generate an image and route it to the best available provider; local reference images are compressed and uploaded.
generate_video writes
Generate video through the MeiGen platform, supporting text-to-video and first-frame image-to-video.

Use cases

Search for design inspiration and prompt templates
Generate product photos, logos, and other images
Animate still images into video
Run local generation with ComfyUI workflows
Generate batch design variations in parallel
Use the standalone CLI in shell scripts or CI

Supported clients

Claude Code
Cursor
Codex
Windsurf
Roo Code
OpenClaw
Hermes Agent

Listed from the project's documentation, not tested by this site.

Overview

MeiGen AI Design MCP turns AI coding tools into design assistants with 8 tools, 1,446 curated prompts, prompt enhancement, image and video generation, ComfyUI workflow management, and parallel batch generation through sub-agents. It supports MeiGen Cloud, OpenAI-compatible APIs, and local ComfyUI backends.

Similar servers

BioMCP 75 · B

One binary. One grammar. Evidence from the biomedical sources you already trust.

★ 659 · Tools 11 Compare with this →
Compare with other MCP servers

Source revision 847fbd995a6a Data synced 2026-10-19 Read the FMRS scoring method