Zonos TTS MCP for Linux
About
Zonos MCP server modified for Linux and GPU optimizations.
Details
- Author
- angrysky56
- Downloads
- 276
- Categories
- Other
Jump to
- Text-to-speech generation directly through Claude.
- Supports four emotions: neutral, happy, sad, angry.
- Multi‑language speech output.
- Audio playback via PulseAudio or PipeWire.
- Optimized for Linux with reduced GPU load.
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Zonos TTS MCP for LinuxCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Install via Smithery (npx -y @smithery/cli install @PhialsBasement/zonos-tts-mcp --client claude) or manually: clone the repo, install dependencies (npm install @modelcontextprotocol/sdk axios zod), build (npm run build), then add the server to Claude’s config under mcpServers. Once configured, Claude can call the speak_response tool with parameters text, language (e.g., "en-us"), and emotion ("neutral", "happy", "sad", "angry").
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"zonos tts mcp for linux": {
"Zonos-TTS-MCP-Linux": {
"command": "npx",
"args": [
"-y",
"@smithery/cli",
"install",
"@PhialsBasement/zonos-tts-mcp",
"--client",
"claude"
]
}
}
}
}McpServers
{
"Zonos-TTS-MCP-Linux": {
"command": "npx",
"args": [
"-y",
"@smithery/cli",
"install",
"@PhialsBasement/zonos-tts-mcp",
"--client",
"claude"
]
}
}Zonos TTS MCP for Linux
A Model Context Protocol integration for Zonos TTS, allowing Claude to generate speech directly on Linux systems.
Overview
This project enables text-to-speech capabilities for Claude AI using the Zonos TTS system. The integration uses the Model Context Protocol (MCP) to allow Claude to generate natural-sounding speech in multiple languages and emotions.
Key Features
- Text-to-speech through Claude
- Multiple emotions support (neutral, happy, sad, angry)
- Multi-language support
- Proper audio playback through PulseAudio/PipeWire
- Compatible with Linux environments
Requirements
- Node.js
- PulseAudio or PipeWire with PulseAudio compatibility layer
- Running instance of Zonos API (PhialsBasement/zonos-api)
- Working audio output device
- Claude Desktop app
Setup
Installing via Smithery
To install the original UNMODIFIED version of Zonos TTS Integration for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @PhialsBasement/zonos-tts-mcp --client claude
Manual installation for Linux with GPU optimizations
1. Make sure you have Zonos API running with the API implementation (PhialsBasement/zonos-api)
2. Clone this repository:
git clone https://github.com/angrysky56/Zonos-TTS-MCP-Linux.git
cd Zonos-TTS-MCP-Linux
3. Install dependencies:
npm install @modelcontextprotocol/sdk axios zod
4. Build the MCP server:
npm run build
5. Add to Claude's config file:
Edit your Claude config file (search for files containing "mcpServers" in your home directory) and add this to the mcpServers section:
"zonos-tts-mcp": {
"command": "node",
"args": [
"/path/to/your/Zonos-TTS-MCP-Linux/dist/server.js"
]
}
Replace /path/to/your/Zonos-TTS-MCP-Linux with the actual path where you installed this project.
Troubleshooting
Error Messages in Console
If you encounter error messages like:
Unexpected token 'P', 'Playing au'... is not valid JSON
This project includes fixes for these common issues:
1. Console output has been redirected to a log file (/tmp/zonos-tts-mcp.log) instead of standard output to prevent JSON parsing errors.
2. The TTS parameters have been optimized for better performance.
Performance Optimization
This fork includes performance optimizations to reduce GPU load:
- Using more efficient sampling parameters (top_p and min_p)
- Improved log handling to prevent interference with MCP communication
Using with Claude
Once configured, Claude can use the speak_response tool:
speak_response(
text="Your text here",
language="en-us", # optional, defaults to en-us
emotion="happy" # optional: "neutral", "happy", "sad", "angry"
)
Credit
This project is a modified version of PhialsBasement/zonos-tts-mcp, adapted for improved Linux compatibility and performance.
The Zonos TTS system is developed by Zyphra and is available through their Zonos API.
License
See the original project for license information.
Notes
- Make sure both the Zonos API server and this MCP server are running
- Audio playback requires proper PulseAudio/PipeWire configuration
- If you experience high GPU usage, you may want to adjust the TTS parameters in the server code
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
