How to Code with Voice in Cursor IDE

Why Local Voice?

Unlike built-in dictate features that might send your voice to a cloud provider, this setup uses Apple MLX Whisper and Kokoro TTS directly on your Mac. It's 100% private, free, and features a hardware-level "Push-to-Talk" lock (the Shift key) so it never accidentally picks up keyboard typing.

Prerequisites

• A Mac with Apple Silicon (M1, M2, M3, or M4).
• Cursor IDE installed.
• Node.js (v18+) installed.

Step 1: Install System Dependencies

Open your terminal and install the required audio libraries using Homebrew:

brew install portaudio espeak-ng

Step 2: Install Voice MCP Server

Install the server globally via NPM so Cursor can easily locate it:

npm install -g voice-mcp-server

Step 3: Connect to Cursor

We need to tell Cursor to use this new capability.

1. Open Cursor and navigate to Settings > Cursor Settings > MCP.
2. Click + Add New MCP Server.
3. Configure it with the following details:
   • Name: voice-mcp-server
   • Type: command
   • Command: voice-mcp-server
4. Click Save. Cursor will connect to the server.

Step 4: Grant Input Monitoring Permissions

To use the Push-to-Talk feature (Shift key) globally, macOS requires permission. Go to System Settings > Privacy & Security > Input Monitoring and toggle Cursor to ON.

Bonus: Need Time to Think?

The Voice MCP Server features an intelligent Standby Mode. If you need a moment, just tell the AI something like "give me a minute" or "let me think." It will gracefully suspend itself, turning off the microphone (and the orange macOS privacy dot) indefinitely. When you are ready to resume, simply press and hold Shift to start talking again!