Voice Mode

v1.0.0Communicationstable

Natural voice conversations with Claude Code

anthropicasrclaudeclaudecodekokoro
Share:
1,190
Stars
0
Downloads
0
Weekly
0/5

What is Voice Mode?

Voice Mode is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to natural voice conversations with claude code

Natural voice conversations with Claude Code

This server falls under the Communication category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Natural voice conversations with Claude Code

Use Cases

Have natural voice conversations with Claude Code.
Use voice input and output for hands-free interaction.
mbailey

Maintainer

LicenseMIT
Languagepython
Versionv1.0.0
UpdatedMay 21, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx voicemode

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Voice Mode

VoiceMode is an MCP server that adds natural voice conversation capabilities to Claude Code and other MCP-compatible clients. It provides speech-to-text using Whisper.cpp (local) or OpenAI Whisper (cloud) and text-to-speech via Kokoro (local) or OpenAI TTS (cloud), enabling hands-free coding sessions, voice-driven code reviews, and conversational pair programming. It supports smart silence detection for natural conversation flow and works on Linux, macOS, Windows (WSL), and NixOS without requiring cloud services if local models are used.

Prerequisites

  • Python 3.10–3.14 installed
  • uv package runner installed (for the recommended install method)
  • Claude Code as your MCP client
  • An OpenAI API key (OPENAI_API_KEY) if using cloud-based speech services; not required if using local Whisper.cpp and Kokoro
  • A microphone and speakers or headphones connected to your machine
1

Install uv if not already installed

VoiceMode's recommended installation uses uv, a fast Python package runner. Install it with the official installer script.

curl -LsSf https://astral.sh/uv/install.sh | sh
2

Install VoiceMode dependencies

Run the voice-mode-install utility via uvx to download and set up the local speech models (Whisper.cpp and Kokoro). This step downloads the model files needed for offline operation.

uvx voice-mode-install
3

Register VoiceMode as an MCP server in Claude Code

Add the VoiceMode MCP server to Claude Code using the claude mcp add command. The --scope user flag makes it available across all your projects.

claude mcp add --scope user voicemode -- uvx --refresh --from voice-mode voicemode-mcp-launcher
4

Set your OpenAI API key (optional, for cloud services)

If you want to use OpenAI Whisper or OpenAI TTS for higher accuracy or faster response, set your API key as an environment variable. Skip this step if you only want to use local models.

export OPENAI_API_KEY="your-openai-api-key"
5

Grant permissions for voice tools in Claude Code

To avoid approval prompts every time voice tools are invoked, add the voicemode tool permissions to your Claude Code settings file.

{
  "permissions": {
    "allow": [
      "mcp__voicemode__converse",
      "mcp__voicemode__service"
    ]
  }
}
6

Start a voice conversation

In Claude Code, use the /voicemode:converse command to start speaking with Claude. The server will listen for your voice, transcribe it, and speak the response back using TTS.

/voicemode:converse

Voice Mode Examples

Client configuration

Claude Desktop configuration for VoiceMode with optional OpenAI API key for cloud speech services.

{
  "mcpServers": {
    "voicemode": {
      "command": "uvx",
      "args": ["--refresh", "--from", "voice-mode", "voicemode-mcp-launcher"],
      "env": {
        "OPENAI_API_KEY": "your-openai-api-key"
      }
    }
  }
}

Prompts to try

These prompts work inside a VoiceMode conversation session in Claude Code.

- "Review the code I just wrote and tell me what could be improved" (spoken aloud)
- "Explain how this function works in simple terms" (spoken aloud)
- "Generate a unit test for the current function and read it back to me"
- "What are the next steps to implement the feature we were just discussing?"

Troubleshooting Voice Mode

No audio input detected during voice conversation

Check that your microphone is set as the default input device in your OS audio settings. On macOS, go to System Settings > Sound > Input. On Linux, check PulseAudio or PipeWire settings with 'pactl list sources'. Run 'voicemode config edit' to verify or change the audio input device.

uvx command not found after installing uv

After running the uv install script, restart your shell or source the profile file shown at the end of the install output (e.g., 'source $HOME/.local/bin/env'). Verify the install with 'which uvx'.

Voice transcription is slow or inaccurate

By default VoiceMode uses local Whisper.cpp models which require CPU resources. For faster, more accurate transcription, set OPENAI_API_KEY and the server will use OpenAI Whisper cloud transcription instead. Alternatively, run 'voicemode config edit' to select a smaller local Whisper model for faster (but less accurate) local transcription.

Frequently Asked Questions about Voice Mode

What is Voice Mode?

Voice Mode is a Model Context Protocol (MCP) server that natural voice conversations with claude code It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Voice Mode?

Follow the installation instructions on the Voice Mode GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Voice Mode?

Voice Mode works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Voice Mode free to use?

Yes, Voice Mode is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

Browse More Communication MCP Servers

Explore all communication servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "voicemode": { "command": "npx", "args": ["-y", "voicemode"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Voice Mode?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides