Chatty MCP

What is Chatty MCP?

Chatty MCP is a voice-enabled server for MCP-supported editors like Cursor and Cline. After every AI request, it speaks an audible summary of the results, letting you know when a task completes and giving you a quick sense of what was produced without reading the output. It is designed for developers who want to multitask while waiting for long AI operations.

How to use Chatty MCP?

Visit the official documentation site for complete installation and configuration steps: https://chatty-mcp.vercel.app/docs/welcome. You can select your preferred text-to-speech engine (system TTS, Kokoro ONNX, or Fish TTS). For Kokoro, you must download the model and place it in the correct directory as described in the Speech Engine documentation.

Key features of Chatty MCP

• Speaks a voice summary after every MCP editor request.
• Supports Cursor, Cline, and any MCP-compatible editor.
• Customizable TTS: system, Kokoro, or Fish voices.
• Enables multitasking—Alt+Tab away while AI thinks.
• Provides quick, audio-based understanding of AI output.

Use cases of Chatty MCP

• Step away from your screen while waiting for a long Cursor request, and return when you hear the announcement.
• Listen to a spoken summary of AI-generated code to decide if adjustments are needed.
• Experience a novel, voice-interactive coding workflow.

FAQ from Chatty MCP

Which editors does Chatty MCP support?

It supports Cursor, Cline, and any MCP-supported editor.

What text-to-speech voices are available?

Chatty MCP can use your system’s built-in TTS, the Kokoro ONNX model for smooth natural voice output, or Fish TTS for anime-style voices.

Is there a documentation website?

Yes. Full documentation, demos, and configuration guides are at https://chatty-mcp.vercel.app/.

Do I need to download any models for voice output?

Only if you choose the Kokoro ONNX engine. You must download the Kokoro model and place it in the correct directory as specified in the Speech Engine documentation.

How does Chatty MCP improve my workflow?

It announces a voice summary after each AI request, so you can multitask (e.g., switch to another window) and quickly assess the AI’s work without reading all the output.