voicemode

voicemode is a local MCP server: Natural voice conversations for AI assistants - STT/TTS via MCP. Install it from pypi as voice-mode.

Claude CodeCodexAntigravity

commandprompt

Terminal

claude mcp add voicemode -e OPENAI_API_KEY=... -e VOICEMODE_DEBUG=... -e VOICEMODE_SKIP_TTS=... -e VOICEMODE_PREFER_LOCAL=... -e VOICEMODE_AUDIO_FORMAT=... -e VOICEMODE_WHISPER_MODEL=... -e VOICEMODE_DISABLE_SILENCE_DETECTION=... -- uvx voice-mode

Run in your project folder. Add --scope user to use it everywhere.

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

{
  "mcpServers": {
    "voicemode": {
      "command": "uvx",
      "args": [
        "voice-mode"
      ],
      "env": {
        "OPENAI_API_KEY": "...",
        "VOICEMODE_DEBUG": "...",
        "VOICEMODE_SKIP_TTS": "...",
        "VOICEMODE_PREFER_LOCAL": "...",
        "VOICEMODE_AUDIO_FORMAT": "...",
        "VOICEMODE_WHISPER_MODEL": "...",
        "VOICEMODE_DISABLE_SILENCE_DETECTION": "..."
      }
    }
  }
}

Fill in the environment variables with your own credentials.

• Runtime: Python with uv (runs with uvx).
• MCP client: Claude Code, Codex, Antigravity, Claude Desktop, Cursor or any other client that speaks MCP over stdio.
• Environment variables: OPENAI_API_KEY, VOICEMODE_DEBUG, VOICEMODE_SKIP_TTS, VOICEMODE_PREFER_LOCAL, VOICEMODE_AUDIO_FORMAT, VOICEMODE_WHISPER_MODEL, VOICEMODE_DISABLE_SILENCE_DETECTION