Labsco
MCP SERVER

MCP MiniMax Music Server

by falahgs

Turn a text prompt into a generated audio track, then poll until the file is ready.

Music & Audio Production
Summary
Generation is two calls, and the key format trips people first.

Submitting and collecting are separate steps — the first returns an ID, the second returns the file — so plan for a wait rather than a single blocking call. The commonest failure is not the model: a 401 almost always means the API key was stored without its `Bearer ` prefix, which the README calls out because the key looks perfectly valid without it.

What it is

A bridge to the MiniMax Music model through the AIML API. You describe the music you want, the server submits the job and returns a generation ID, then a second call with that ID reports status and hands back the finished audio file URL.

What you get
  • `generate_audio` takes a `prompt` describing the music you want
  • An optional `model` argument, set to `minimax-music`
  • An optional `reference_audio_url` for style matching
  • An optional `generation_id` — pass the ID from a previous run to check on it instead of starting a new one
  • The completed response carries the audio `url`, `content_type`, `file_name` and `file_size`
  • Lyrics are written between `##` delimiters in the prompt
Requirements

Node.js v16 or higher and TypeScript v5.3.3 or higher. An AIML API key is required, and its format matters: the value must carry the `Bearer ` prefix, spelled `"AIML_API_KEY": "Bearer your-aiml-api-key-here"` in the client's `env` block. Published as mcp-minimax-music-server version 0.1.0; the documented path is clone, `npm install`, `npm run build`, then point your client at `build/index.js`. Windows paths in the config need double backslashes.