MCP.Pizza Chef: MiniMax-AI
MiniMax's voice, picture, video and music generators arrive as ten tools your assistant can call: read text aloud, clone a voice from a recording, invent a new voice from a description, make images, turn text or a photo into video, compose music, check a video job, list voices and play a finished file. Every generation spends real money from a paid MiniMax account, voice cloning needs identity verification first, and the project has had no updates since July 2025.
Unmaintained · No commits in 14 months.
Images/Design
Other
Use This MCP server To
Turn a written script into narration in a chosen voice Generate a short video from a single still photo Create cover art from a written description Compose background music for a clip I am editing Design a brand-new voice from a description Save generated audio straight to a folder I pick
README
MiniMax-MCP-JS FAQ
Is it still being updated?
Not lately. The last change landed in July 2025, so it has been quiet for around a year. It is not archived and the company still publishes a sister version in another programming language.
Does this cost money?
Yes. Nearly every tool bills a paid MiniMax account with a balance, and the tools themselves carry a warning to use them only when you explicitly ask.
Do I need a key?
Yes, from MiniMax — and it must be paired with the matching regional address for global or mainland China accounts, or you get an invalid-key error. The project's setup notes contradict themselves on which address belongs to which region, so copy the one from its region table.
Can I use this to clone somebody's voice?
Technically yes, but MiniMax blocks cloning until you complete real-name identity verification, and you should only clone a voice you have permission to use.
Which apps does it work in?
Claude Desktop and Cursor have documented setups, plus a one-command Smithery install, and it works in other assistants that support MCP.
How many tools does it add?
Ten, covering speech, voice cloning, voice design, images, two kinds of video, music, job status, voice listing and audio playback.
Where do the finished files go?
Into a folder you choose in the setup, defaulting to your desktop — or you can have links handed back instead of files.
Can I browse the available voices?
Only on the mainland China address. The voice-listing tool is unavailable on the global one, so global users have to know voice names in advance.
How hard is setup?
You paste a short block into your assistant's settings with your key, your chosen region address and an output folder.