PluginBench
MCP Server
Active
MIT

dev.voicemode/voicemode MCP Server

dev.voicemode/voicemode

Natural voice conversations with Claude Code—speak naturally, hear responses immediately with STT/TTS via MCP.

RepositoryRegistry
How to choose

What this kind of tool is for

  • Draft, send, or triage messages and email from inside your agent.
  • Pipe notifications from a build or alert into your team channel automatically.
  • Summarize a busy channel thread into the handful of points that matter.

How to compare servers like this

Channel coverage:
Check the specific app (email, chat, notifications) and whether it reads as well as writes.
Scope of access:
Messages are sensitive; review what the tool can see before connecting an account.
Rate behavior:
Bulk-sending tools should respect provider limits — avoid anything that floods.
Maintenance signal:
APIs in this space change often; recent updates are a strong signal.

This page carries your copy-paste config: pick your MCP client in the install section on this page and use the copy button.

What is the dev.voicemode/voicemode MCP server?

VoiceMode is an MCP server that enables natural voice conversations with Claude Code and other MCP-capable agents through speech-to-text and text-to-speech. It supports both cloud-based services (OpenAI) and local offline alternatives (Whisper.cpp, Kokoro) for privacy-conscious users. Perfect for hands-free interaction when typing or looking at screens isn't practical.

VoiceMode lets you have natural voice conversations with Claude Code using your microphone and speakers. It handles speech-to-text transcription and text-to-speech responses with smart silence detection and low latency. You can run it entirely offline with local voice services or use cloud APIs—it works seamlessly across Linux, macOS, Windows, and NixOS.

How to install dev.voicemode/voicemode

Copy-paste configuration for popular MCP clients.

transport: stdio
Config generated by PluginBench — verify against the source before use.
Environment / auth
  • OPENAI_API_KEY
    secret

    OpenAI API key for cloud-based STT/TTS (optional - local services can be installed)

  • VOICEMODE_DEBUG

    Enable debug mode with detailed logging (true/false)

  • VOICEMODE_SKIP_TTS

    Skip TTS and show text only for faster response (true/false)

  • VOICEMODE_PREFER_LOCAL

    Prefer local services over cloud when available (true/false, default: true)

  • VOICEMODE_AUDIO_FORMAT

    Audio format: pcm, mp3, wav, flac, aac, opus (default: pcm)

  • VOICEMODE_WHISPER_MODEL

    Whisper model: tiny, base, small, medium, large (default: base)

  • VOICEMODE_DISABLE_SILENCE_DETECTION

    Disable silence detection for continuous recording (true/false)

~/Library/Application Support/Claude/claude_desktop_config.json
JSON config
{
  "mcpServers": {
    "voicemode": {
      "command": "uvx",
      "args": [
        "voice-mode"
      ],
      "env": {
        "OPENAI_API_KEY": "<YOUR_OPENAI_API_KEY>",
        "VOICEMODE_DEBUG": "<YOUR_VOICEMODE_DEBUG>",
        "VOICEMODE_SKIP_TTS": "<YOUR_VOICEMODE_SKIP_TTS>",
        "VOICEMODE_PREFER_LOCAL": "<YOUR_VOICEMODE_PREFER_LOCAL>",
        "VOICEMODE_AUDIO_FORMAT": "<YOUR_VOICEMODE_AUDIO_FORMAT>",
        "VOICEMODE_WHISPER_MODEL": "<YOUR_VOICEMODE_WHISPER_MODEL>",
        "VOICEMODE_DISABLE_SILENCE_DETECTION": "<YOUR_VOICEMODE_DISABLE_SILENCE_DETECTION>"
      }
    }
  }
}

Tools & capabilities

Tools this server exposes to the agent.

  • converse — Start a natural voice conversation with Claude Code, with automatic speech-to-text transcription and text-to-speech responses
  • service — Manage local voice services (Whisper.cpp for STT, Kokoro for TTS) or configure cloud service integration

Use cases

  • Have hands-free conversations with Claude while walking, cooking, or doing other activities
  • Debug code and get assistance without typing, using only voice commands and audio responses
  • Use local speech services to keep conversations completely private and offline
  • Switch seamlessly between cloud-based and local voice services depending on your privacy needs
  • Get immediate audio feedback from Claude Code with natural conversation flow and smart silence detection

dev.voicemode/voicemode MCP server FAQ

What is VoiceMode?

VoiceMode is an MCP server that adds natural voice conversation capabilities to Claude Code. You speak into your microphone, it transcribes your speech, sends it to Claude, and reads Claude's response aloud—all with low latency and smart silence detection.

Is VoiceMode free?

VoiceMode itself is free and open-source (MIT license). If you use cloud services like OpenAI's Whisper or TTS, you'll incur API costs. Local alternatives (Whisper.cpp, Kokoro) are free and run entirely on your machine.

How do I install VoiceMode in Claude Code?

The fastest way is via the Claude Code plugin marketplace: `claude plugin marketplace add mbailey/voicemode`, then `claude plugin install voicemode@voicemode`. Alternatively, install the Python package (`uvx voice-mode-install`) and add it as an MCP server.

Can I use VoiceMode offline?

Yes. VoiceMode supports local speech services: Whisper.cpp for speech-to-text and Kokoro for text-to-speech. Both run entirely on your machine without requiring internet or API keys.

What platforms does VoiceMode support?

VoiceMode runs on Linux, macOS, Windows (native or WSL), and NixOS. It requires Python 3.10-3.14 and a computer with a microphone and speakers.

Do I need to set up permissions?

VoiceMode works out of the box. For seamless operation without permission prompts, you can optionally add permissions to `~/.claude/settings.json` to allow the `mcp__voicemode__converse` and `mcp__voicemode__service` tools.

README (reference)

Source of truth, from the repository.

VoiceMode

Natural voice conversations with Claude Code (and other MCP capable agents)

PyPI Downloads PyPI Downloads PyPI Downloads

VoiceMode enables natural voice conversations with Claude Code. Voice isn't about replacing typing - it's about being available when typing isn't.

Perfect for:

  • Walking to your next meeting
  • Cooking while debugging
  • Giving your eyes a break after hours of screen time
  • Holding a coffee (or a dog)
  • Any moment when your hands or eyes are busy

See It In Action

VoiceMode Demo

Quick Start

Requirements: Computer with microphone and speakers

Option 1: Claude Code Plugin (Recommended)

The fastest way for Claude Code users to get started:

# Add the VoiceMode marketplace
claude plugin marketplace add mbailey/voicemode

# Install VoiceMode plugin
claude plugin install voicemode@voicemode

## Install dependencies (CLI, Local Voice Services)

/voicemode:install

# Start talking!
/voicemode:converse

Option 2: Python installer package

Installs dependencies and the VoiceMode Python package.

# Install UV package manager (if needed)
curl -LsSf https://astral.sh/uv/install.sh | sh

# Run the installer (sets up dependencies and local voice services)
uvx voice-mode-install

# Add to Claude Code
claude mcp add --scope user voicemode -- uvx --refresh --from voice-mode voicemode-mcp-launcher

# Optional: Add OpenAI API key as fallback for local services
export OPENAI_API_KEY=your-openai-key

# Start a conversation
claude converse

For manual setup, see the Getting Started Guide.

Features

  • Natural conversations - speak naturally, hear responses immediately
  • Works offline - optional local voice services (Whisper STT, Kokoro TTS)
  • Low latency - fast enough to feel like a real conversation
  • Smart silence detection - stops recording when you stop speaking
  • Privacy options - run entirely locally or use cloud services

Compatibility

Platforms: Linux, macOS, Windows (native or WSL), NixOS Python: 3.10-3.14

Configuration

VoiceMode works out of the box. For customization:

# Set OpenAI API key (if using cloud services)
export OPENAI_API_KEY="your-key"

# Or configure via file
voicemode config edit

See the Configuration Guide for all options.

Permissions Setup (Optional)

To use VoiceMode without permission prompts, add to ~/.claude/settings.json:

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

See the Permissions Guide for more options.

Local Voice Services

For privacy or offline use, install local speech services:

  • Whisper.cpp - Local speech-to-text
  • Kokoro - Local text-to-speech with multiple voices

These provide the same API as OpenAI, so VoiceMode switches seamlessly between them.

Installation Details

<details> <summary><strong>System Dependencies by Platform</strong></summary>

Ubuntu/Debian

sudo apt update
sudo apt install -y ffmpeg gcc libasound2-dev libasound2-plugins libportaudio2 portaudio19-dev pulseaudio pulseaudio-utils python3-dev

WSL2 users: The pulseaudio packages above are required for microphone access.

Fedora/RHEL

sudo dnf install alsa-lib-devel ffmpeg gcc portaudio portaudio-devel python3-devel

macOS

brew install ffmpeg node portaudio

NixOS

# Use development shell
nix develop github:mbailey/voicemode

# Or install system-wide
nix profile install github:mbailey/voicemode
</details> <details> <summary><strong>Alternative Installation Methods</strong></summary>

From source

git clone https://github.com/mbailey/voicemode.git
cd voicemode
uv tool install -e .

NixOS system-wide

# In /etc/nixos/configuration.nix
environment.systemPackages = [
  (builtins.getFlake "github:mbailey/voicemode").packages.${pkgs.system}.default
];
</details>

Troubleshooting

ProblemSolution
No microphone accessCheck terminal/app permissions. WSL2 needs pulseaudio packages.
UV not foundRun curl -LsSf https://astral.sh/uv/install.sh | sh
OpenAI API errorVerify OPENAI_API_KEY is set correctly
No audio outputCheck system audio settings and available devices

Save Audio for Debugging

export VOICEMODE_SAVE_AUDIO=true
# Files saved to ~/.voicemode/audio/YYYY/MM/

Documentation

Full documentation: voicemode.dev

Links

License

MIT - A Failmode Project


mcp-name: dev.voicemode/voicemode

Related MCP servers

MCP server for local speech translation (EN ↔ 中文) via Whisper + Claude + Piper

3
TypeScript
MIT
View repository →

Coordination hub for AI coding agents: message teammates, ask humans, audit every event.

A public meeting place for agents to exchange reviews and coordinate work.

Send email through your own Amazon SES account, and check domains, suppressions and sandbox status.

1
TypeScript
MIT
View repository →

Read-only production error evidence for coding agents investigating failures.

Create, update, and revoke Apple Wallet and Google Wallet passes from any MCP client.