Brainiall Speech AI - MCP Server

MCP server providing AI-powered pronunciation assessment, speech-to-text, and text-to-speech tools via Brainiall Speech AI APIs.

Tools

Tool	Description
`assess_pronunciation`	Score how accurately a speaker pronounced a given text (0-100), with per-word and phoneme-level feedback
`transcribe_speech`	Transcribe speech audio into text with automatic language detection
`synthesize_speech`	Convert text to natural-sounding speech audio (MP3)
`list_voices`	List all available text-to-speech voices

Prerequisites

Get your free API key at app.brainiall.com.

Installation

Using pip

pip install fastmcp httpx

Using Docker

docker build -t brainiall-mcp .
docker run -e BRAINIALL_API_KEY=your-key -p 8000:8000 brainiall-mcp

Configuration

Set the BRAINIALL_API_KEY environment variable:

export BRAINIALL_API_KEY=your-api-key

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "brainiall-speech-ai": {
      "command": "fastmcp",
      "args": ["run", "server.py"],
      "env": {
        "BRAINIALL_API_KEY": "your-api-key"
      }
    }
  }
}

Cursor / VS Code

Add to your MCP settings:

{
  "mcpServers": {
    "brainiall-speech-ai": {
      "command": "fastmcp",
      "args": ["run", "/path/to/server.py"],
      "env": {
        "BRAINIALL_API_KEY": "your-api-key"
      }
    }
  }
}

Streamable HTTP (Docker / Remote)

docker run -e BRAINIALL_API_KEY=your-key -p 8000:8000 brainiall-mcp

Connect your MCP client to http://localhost:8000/mcp.

Running Locally

# stdio (default for Claude Desktop / Cursor)
BRAINIALL_API_KEY=your-key fastmcp run server.py

# streamable-http
BRAINIALL_API_KEY=your-key fastmcp run server.py --transport streamable-http --port 8000

API Reference

All tools communicate with the Brainiall Speech AI API at https://api.brainiall.com. Full API documentation is available at app.brainiall.com.

assess_pronunciation

Parameter	Type	Required	Description
`text`	string	Yes	Reference text the user should have read
`audio_base64`	string	Yes	Base64-encoded audio (WAV or MP3)
`language`	string	No	Language code, default `en-US`

transcribe_speech

Parameter	Type	Required	Description
`audio_base64`	string	Yes	Base64-encoded audio (WAV, MP3, WEBM, OGG)
`language`	string	No	Language hint for transcription

synthesize_speech

Parameter	Type	Required	Description
`text`	string	Yes	Text to convert to speech
`voice`	string	No	Voice ID, default `alloy`
`speed`	float	No	Speed multiplier (0.5-2.0), default `1.0`

list_voices

No parameters. Returns available voice options with IDs and supported languages.

License

MIT

brainiall-mcp-server