README.md
Minimal cross-platform MCP server for text-to-speech using Microsoft Edge TTS.
mcp-edge-tts - Give Claude a Voice!
Minimal cross-platform MCP server for text-to-speech using Microsoft Edge TTS.
Let Claude speak! Works with Claude Code and Claude Desktop.
Слава Україні!
Features
- 🎙️ 300+ voices in 50+ languages
- 🖥️ Cross-platform: Windows, macOS, Linux
- 🔑 Zero API keys required
- ⚙️ Customizable: speed, volume, pitch
- 🤖 Works with Claude Code and Claude Desktop
Installation
Via pip (recommended)
pip install mcp-edge-tts
From source
Windows
git clone https://github.com/s-n-n/edge-tts-mcp.git
cd edge-tts-mcp
python -m venv venv
venv\Scripts\activate
pip install -e .
macOS / Linux
git clone https://github.com/s-n-n/edge-tts-mcp.git
cd edge-tts-mcp
python3 -m venv venv
source venv/bin/activate
pip install -e .
Setup
Add to your MCP configuration:
If installed via pip
{
"mcpServers": {
"edge-tts": {
"command": "python",
"args": ["-m", "mcp_edge_tts"]
}
}
}
If installed from source
Claude Code (`.mcp.json`)
Windows:
{
"mcpServers": {
"edge-tts": {
"command": "C:\\path\\to\\edge-tts-mcp\\venv\\Scripts\\python.exe",
"args": ["-m", "mcp_edge_tts"]
}
}
}
macOS / Linux:
{
"mcpServers": {
"edge-tts": {
"command": "/path/to/edge-tts-mcp/venv/bin/python",
"args": ["-m", "mcp_edge_tts"]
}
}
}
Usage
speak
Speak text aloud.
speak("Hello, world!")
speak("Привіт!", voice="uk-UA-PolinaNeural")
speak("Fast!", rate="+50%")
Parameters:
| Parameter | Description | Example |
|---|---|---|
text |
Text to speak | "Hello" |
voice |
Voice name | "en-US-AriaNeural" |
rate |
Speed (-50% to +100%) | "+20%" |
volume |
Volume (-50% to +100%) | "+10%" |
pitch |
Pitch (-50Hz to +50Hz) | "+5Hz" |
list_available_voices
List voices, optionally filtered by language.
list_available_voices()
list_available_voices("uk")
get_config
Show current settings and available audio players.
Configuration
Set defaults via environment variables or .env file:
EDGE_TTS_VOICE=uk-UA-OstapNeural
EDGE_TTS_RATE=+0%
EDGE_TTS_VOLUME=+0%
EDGE_TTS_PITCH=+0Hz
EDGE_TTS_PLAYER=auto
Audio Players
Auto-detected by platform:
| Platform | Default Player | Alternatives |
|---|---|---|
| Windows | PowerShell MediaPlayer (built-in) | ffplay, mpv |
| macOS | afplay (built-in) | ffplay, mpv |
| Linux | ffplay | mpv, paplay |
Popular Voices
| Language | Voice | Gender |
|---|---|---|
| Ukrainian | uk-UA-OstapNeural | Male |
| Ukrainian | uk-UA-PolinaNeural | Female |
| English (US) | en-US-AriaNeural | Female |
| English (US) | en-US-GuyNeural | Male |
| German | de-DE-ConradNeural | Male |
| French | fr-FR-DeniseNeural | Female |
| Spanish | es-ES-AlvaroNeural | Male |
| Chinese | zh-CN-XiaoxiaoNeural | Female |
Run list_available_voices() for full list.
Requirements
- Python 3.10+
- No additional software on Windows/macOS (uses built-in players)
- Linux:
ffmpegormpvrecommended
License
MIT
Made with ❤️ in Ukraine
Tools (3)
speakSpeak text aloud using Microsoft Edge TTS.list_available_voicesList voices, optionally filtered by language.get_configShow current settings and available audio players.Environment Variables
EDGE_TTS_VOICEDefault voice to use for speechEDGE_TTS_RATEDefault speech rate (-50% to +100%)EDGE_TTS_VOLUMEDefault volume (-50% to +100%)EDGE_TTS_PITCHDefault pitch (-50Hz to +50Hz)EDGE_TTS_PLAYERAudio player to useConfiguration
claude_desktop_config.json
{"mcpServers": {"edge-tts": {"command": "python", "args": ["-m", "mcp_edge_tts"]}}}Try it
→Speak the following text aloud: 'Hello, I am Claude, your AI assistant.'
→List all available Ukrainian voices.
→Read this summary using the uk-UA-PolinaNeural voice at a slightly faster rate.
→What are my current Edge TTS configuration settings?