Overview
What is Zonos MCP Integration?
A Model Context Protocol integration for Zonos TTS that allows Claude to generate speech directly. It acts as a bridge between an AI assistant (Claude) and a running Zonos API instance, enabling text-to-speech playback on the user's machine.
How to use Zonos MCP Integration?
Install via Smithery (npx -y @smithery/cli install @PhialsBasement/zonos-tts-mcp --client claude) or manually by cloning the repo and adding the server to Claude's MCP config. Once configured, Claude automatically exposes a speak_response tool that accepts text (required), language (optional, defaults to en-us), and emotion (optional: neutral, happy, sad, angry).
Key features of Zonos MCP Integration
- Text-to-speech generation through Claude
- Multiple emotion support (neutral, happy, sad, angry)
- Multi-language support
- Proper audio playback via PulseAudio
Use cases of Zonos MCP Integration
- Giving Claude a spoken voice for interactive demos or voice assistants
- Generating audio narration from Claude's text responses
- Adding emotion and language variety to AI-generated speech
FAQ from Zonos MCP Integration
What are the system requirements?
Node.js, a running instance of Zonos API (PhialsBasement/zonos-api), PulseAudio properly configured, and a working audio output device.
How do I manually install Zonos MCP Integration?
Clone the repo, install dependencies (npm install @modelcontextprotocol/sdk axios), build (npm run build), and add the server to Claude's config file under mcpServers with command node and args pointing to dist/server.js.
How do I use it with Claude?
After configuration, Claude automatically knows how to use the speak_response tool. Provide the text you want spoken, optionally specify language and emotion.
Does Zonos MCP Integration work without a Zonos API server?
No, the MCP server requires a running Zonos API instance (PhialsBasement/zonos-api) to function. Both servers must be running simultaneously.
