MusicGPT

You run a MusicGPT MCP Server to enable AI-powered music generation, voice transformation, and audio processing through MCP clients. This server exposes a wide range of tools that let you generate music, transform voices, manipulate audio, and analyze audio content, all accessible via MCP-compatible clients with a simple local or remote setup.

How to use

Set up the server locally, start it, and connect MCP clients to perform tasks such as generating music from prompts, creating cover songs, applying voice changes, or extracting stems from audio.

How to install

Prerequisites you need before installing include Node.js and npm. You will install the MCP server package, or you can run it directly from source.

npm install mcp-server-musicgpt

or install from source:

git clone https://github.com/pasie15/mcp-server-musicgpt.git
cd mcp-server-musicgpt
npm install
npm run build

Finally, set your MusicGPT API key in the environment before starting the server.

## Configuration

Configure the client and server with your MusicGPT API key. The server uses this key to access MusicGPT services.

export MUSICGPT_API_KEY="your_api_key_here"

Optional settings (adjust as needed):

export MUSICGPT_BASE_URL="https://api.musicgpt.com/api/public/v1" export MUSICGPT_TIMEOUT="60000"

These environment variables are used by the server to authorize and connect to MusicGPT services.

Connecting a client via MCP (example configurations)

You can connect a client by adding an MCP server entry to the client configuration. The following examples show how to run the server with npx and how to reference the API key.

{
  "mcpServers": {
    "musicgpt": {
      "command": "npx",
      "args": ["-y", "mcp-server-musicgpt"],
      "env": {
        "MUSICGPT_API_KEY": "your_api_key_here"
      }
    }
  }
}

Alternative (install from source) for MacOS/Linux:

{
  "musicgpt": {
    "command": "node",
    "args": ["/path/to/mcp-server-musicgpt/dist/index.js"],
    "env": {"MUSICGPT_API_KEY": "your_api_key_here"}
  }
}

Alternative (Windows):

{
  "musicgpt": {
    "command": "node",
    "args": ["C:\\path\\to\\mcp-server-musicgpt\\dist\\index.js"],
    "env": {"MUSICGPT_API_KEY": "your_api_key_here"}
  }
}

These configurations start the MCP server locally and connect it to your MusicGPT account.

## Available tools

### get\_conversion\_by\_id

Check the status of a conversion task by task\_id or conversion\_id and retrieve results when available.

### get\_all\_voices

List all available AI voices with pagination controls.

### search\_voices

Find voices by name or partial matches.

### generate\_music

Create music from a text prompt with optional lyrics and voice settings.

### create\_cover\_song

Produce a cover by applying a selected voice to an existing audio track.

### generate\_sound\_effect

Generate sound effects from descriptive prompts.

### generate\_lyrics

Generate song lyrics based on a theme or prompt.

### voice\_changer

Apply AI voice transformations to an input audio file.

### text\_to\_speech

Convert text to natural-sounding speech using AI voices.

### extract\_audio

Isolate vocals, instruments, or specific stems from a track.

### denoise\_audio

Remove background noise from audio.

### deecho\_audio

Remove echo from audio recordings.

### dereverb\_audio

Remove reverb from audio.

### convert\_audio\_format

Convert audio between formats such as mp3, wav, flac, ogg, m4a.

### cut\_audio

Trim audio to a specified time range.

### change\_audio\_speed

Change the playback speed of audio.

### master\_audio

Apply professional mastering to audio.

### remix\_audio

Create remixes in various styles.

### extend\_audio

Extend audio with AI-based continuation.

### inpaint\_audio

Fill gaps or repair sections in audio.

### sing\_over\_instrumental

Add AI vocals to an instrumental track.

### transcribe\_audio

Transcribe speech to text from audio.

### extract\_key\_bpm

Detect key and tempo from audio.

### audio\_to\_midi

Convert audio to MIDI format.