mcpdirectory.dev
9,129 servers · updated 23 Sept

Voicemode MCP server

community Active last commit 2 d
{
  "mcpServers": {
    "voicemode": {
      "command": "uvx",
      "args": [
        "voice-mode"
      ],
      "env": {
        "OPENAI_API_KEY": "sk-..."
      }
    }
  }
}
file .mcp.json (in your project root)install uvx voice-modeversion 8.12.0

What it does

VoiceMode enables natural voice conversations with Claude and other MCP-capable agents using speech-to-text and text-to-speech. It supports both cloud-based services (OpenAI) and local, offline alternatives (Whisper.cpp and Kokoro) for privacy-conscious users.

  • Interact with Claude Code through voice while walking, cooking, or doing hands-on work where typing is impractical.
  • Run speech recognition and text-to-speech entirely locally for offline conversations or privacy requirements.
  • Debug code or ask questions verbally while maintaining eye contact or keeping hands free for other tasks.

Tools this server exposes

ToolDoes
converseStart a natural voice conversation with Claude or another MCP agent, with automatic silence detection and immediate spoken responses.
serviceManage voice services configuration, including speech-to-text and text-to-speech provider selection.

Before you enable it

An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.