dev.voicemode/voicemodemedia

voicemode

Verified · yesterday

Natural voice conversations for AI assistants - STT/TTS via MCP

Install

claude mcp add voicemode -- uvx voice-mode

Requires: OPENAI_API_KEY, VOICEMODE_DEBUG, VOICEMODE_SKIP_TTS, VOICEMODE_PREFER_LOCAL, VOICEMODE_AUDIO_FORMAT, VOICEMODE_WHISPER_MODEL, VOICEMODE_DISABLE_SILENCE_DETECTION

Our take

Adds spoken conversation to Claude Code and other agents — local Whisper and Kokoro for the privacy-conscious, OpenAI voices if you don't care. The local-first option is the differentiator; your audio never has to leave the machine. Setup involves more moving parts than a typical server.

reviewed by hand · 2026-07-05

Something wrong or dead here? Report it

Verification record

last verified: yesterday
github stars: 1.3k
last commit: 7 days ago
archived: no
license: MIT
in registry since: 2026-06-26

github.com/mbailey/voicemode