Skip to content
MCP Shelf

Text-to-speech, voice cloning, images and video from MiniMax.

  • Started in our test
  • ★ 1.6K GitHub stars
  • 2K downloads a week
  • Last commit 20 Aug 2026

At a glance

MiniMax's own server wraps its media models: speech in many voices and languages, voice cloning from a sample, image generation and short video clips. Files are saved to a folder you choose.

Note the API host: accounts from the global site and from mainland China use different hosts, and a mismatch is the usual reason it fails to authenticate.

Made by
MiniMax
Licence
MIT
Latest version
0.0.19 (21 Aug 2026)
PyPI package
minimax-mcp
GitHub stars
1.6K
Downloads
2K a week
Last commit
20 Aug 2026
Tools
8
Runs on
Your computer

Our check: We ran this exact config on 26 Oct 2026: it started and answered with 8 tools. The npm package called "minimax-mcp-server" comes from a different repository (mgolec/minimax-integration). MiniMax publishes this server to PyPI as minimax-mcp.

What's in the pod: its tools

We found 8 tools when we started it on 26 Oct 2026. 2 only read. The rest can change things, so keep your app's "ask before each tool call" setting on for them.

Changes things6 tools · Can create, edit, send or delete.

ToolWhat it is for
generate_videoMakes a short video clip from a text prompt, optionally starting from an image, with the Hailuo models.
play_audioPlays an audio file through your speakers.
text_to_audioReads text aloud in a chosen voice, speed and emotion and saves the audio.
text_to_imageCreates images from a prompt in the aspect ratio you ask for.
voice_cloneBuilds a reusable voice from a sample recording.
voice_designInvents a new voice from a written description and gives you a preview.

Only reads2 tools · Looks things up and changes nothing.

ToolWhat it is for
list_voicesLists the voices you can use for speech, including any you have cloned.
query_video_generationChecks whether a video job has finished and fetches the result.

Grouped and described by us. The README on GitHub lists the inputs each tool takes.

Set it up

Pick your app for the exact snippet and where it goes.

Where it goes: claude_desktop_config.json (Settings → Developer → Edit Config)

Claude Desktop
{
  "mcpServers": {
    "minimax": {
      "command": "uvx",
      "args": [
        "minimax-mcp"
      ],
      "env": {
        "MINIMAX_API_KEY": "your-key",
        "MINIMAX_API_HOST": "https://api.minimax.io"
      }
    }
  }
}

New to this? See the setup guides for Claude Code, Cursor, Claude Desktop and ChatGPT.

Settings

Names are exactly as the project spells them. Required ones are marked.

SettingWhat it does
MINIMAX_API_KEYEnvironment variable · requiredYour MiniMax API key. It must come from the same region as the host below.
MINIMAX_API_HOSTEnvironment variable · requiredWhich MiniMax region to call: the global platform or mainland China.https://api.minimax.io
MINIMAX_MCP_BASE_PATHEnvironment variableFolder where downloaded audio, images and video are saved./Users/you/Desktop
MINIMAX_API_RESOURCE_MODEEnvironment variableurl (default) hands back links to the files; local downloads them to the base path.local

What it can touch

It can read

  • Text prompts, and audio or image files you pass in for cloning or video

It can change

  • Saves generated audio, images and video to your output folder in local mode
  • Spends credit on your MiniMax account for every generation

Safer setup: Use url mode so nothing is written to disk, and watch your MiniMax balance, since video is the priciest call.

Before you start

  • A MiniMax API key
  • uv (for uvx)

If it does not work

API Error: invalid api key
The key and host are from different regions. A key from minimax.io needs https://api.minimax.io; one from minimaxi.com needs https://api.minimaxi.com.
spawn uvx ENOENT
Put the full path from which uvx in the command field so your desktop app can find it.
Claude Desktop on Windows does not load it
Turn on Developer Mode from the Help menu in Claude Desktop, then restart.
Video calls seem to hang
Video takes minutes. Ask for async mode and check back with query_video_generation.

Peas in a pod: similar servers

How MiniMax lines up against other design, 3d and games servers on the shelf.

ServerToolsRuns onLicenceStars
MiniMax8Your computerMIT1.6K
Blender MCP9Your computerMIT32.4K
Framelink Figma MCP2Your computerMIT16.1K
Talk to Figma40Your computer–7.1K
Godot MCP14Your computerMIT6.4K

Questions

What can it make?

Speech in many voices, cloned or designed voices, images and short videos up to 10 seconds at 1080p.

Is it free?

The server is MIT licensed, but every generation is billed to your MiniMax account.

Does it need Python?

It runs with uvx, so you need uv installed. A JavaScript version also exists from the same company.

Full docs on GitHubMiniMax-AI/MiniMax-MCP

Something wrong on this page? Tell us.

More in Design, 3D and games