Edge-TTS MCP Server

A Model Context Protocol (MCP) server that provides speech synthesis services for AI agents that leverage the text-to-speech capabilities of Microsoft Edge.

overview

This MCP server uses the edge-tts library to provide text-to-speech capabilities, and is designed as a tool to enable AI agents to respond in a natural voice.

Related MCP server: AivisSpeech MCP Server

function

• Text to speech conversion

• Multiple voice and language support

• Adjust audio speed and pitch

• Streaming audio data

install

pip install "edge_tts_mcp_server"

Or if you want to install in development mode:

git clone https://github.com/yuiseki/edge_tts_mcp_server.git
cd edge_tts_mcp_server
pip install -e .

How to use

Example setup in VS Code

Example of setting in VS Code settings.json:

"mcp": {
  "servers": {
    "edge-tts": {
      "command": "uv",
      "args": [
        "--directory",
        "C:\\Users\\__username__\\src\\edge_tts_mcp_server\\src\\edge_tts_mcp_server",
        "run",
        "server.py"
      ]
    }
  }
}

Use with MCP Inspector

Runs as a standard MCP server:

mcp dev server.py

Running with uvx (uvicorn)

If you run it as a FastAPI based server under uv:

uv --directory path/to/edge_tts_mcp_server/src/edge_tts_mcp_server run server.py

Command line options:

edge-tts-mcp --host 0.0.0.0 --port 8080 --reload

API endpoint

When running in FastAPI mode, the following endpoints are available:

• / - API information

• /health - health check

• /voices - List of available voices (optionally filterable by ?locale=ja-JP , etc.)

• /mcp - MCP API endpoint

license

MIT