Edge TTS MCP

A Model Context Protocol (MCP) server for Microsoft Edge Text-to-Speech service that allows AI assistants to read text aloud with natural-sounding voices.

Language

• 简体中文

• 日本語

Features

• Generate lifelike speech from text input

• Support for multiple voice options

• Customizable speech parameters (rate, volume, pitch)

• Optional audio saving capability

• Easy integration with Cline and other MCP-compatible clients

Installation

Prerequisites

• Node.js (v16 or later)

• Bun (v1.0.0 or later)

Setup

1. Clone the repository:

2. Install dependencies:

3. Build the project:

Configuration

Environment Variables

The Edge TTS MCP server supports the following environment variables:

You can set these environment variables before starting the server.

Usage

Starting the Server

Integrating with Cline

To use this MCP server with Cline, add the following configuration to your Cline config:

Replace /path/to/edge-tts-mcp with the actual path to your installation.

MCP Tool Parameters

The MCP server exposes the following tool:

Tool Name: speech_text_aloud

Parameters:

• input (string): The text to be converted to speech and read aloud

Using with Node.js

You can also run the server using Node.js instead of Bun:

For Cline integration with Node.js, update your configuration:

Audio File Storage

When the SAVE_AUDIO environment variable is set to true, audio files will be saved in the dist directory by default. Each file is named with a random UUID to prevent overwriting.

Configuration for Other Clients

5ire or Claude

You can also configure this MCP server in other clients like 5ire or Claude. Here's an example configuration:

⚠️ Path Format Warning: Pay attention to the path format in your configuration:

• Windows: Uses backslashes (\) and needs to be escaped in JSON as \\ or converted to forward slashes (/)

• macOS/Linux: Uses forward slashes (/)

Examples:

• Windows path: C:\\Users\\username\\edge-tts-mcp\\dist\\index.js or C:/Users/username/edge-tts-mcp/dist/index.js

• macOS/Linux path: /Users/username/edge-tts-mcp/dist/index.js

Incorrect path formatting is a common cause of setup issues across different operating systems.

Make sure to adjust the file path according to your actual installation directory.

⚠️ Important Notice: Currently, there are known issues with MCP integration in Cherry-Studio. The configuration above may not work properly in Cherry-Studio. We recommend using Cline or other well-tested MCP clients until these issues are resolved.

Example Usage in an AI Assistant

When your AI assistant needs to read text aloud, it can use a prompt like:

The assistant will call the speech_text_aloud tool with the appropriate input text.

Voice Options

Microsoft Edge TTS provides many voices across different languages. Some popular options include:

• en-US-AriaNeural (English, US, Female)

• en-US-GuyNeural (English, US, Male)

• zh-CN-XiaoxiaoNeural (Chinese, Female)

• ja-JP-NanamiNeural (Japanese, Female)

• de-DE-KatjaNeural (German, Female)

• fr-FR-DeniseNeural (French, Female)

For a complete list of available voices, refer to the Microsoft Edge TTS documentation.

License

MIT