/voicemode:install

Install VoiceMode and all dependencies needed for voice conversations.

Quick Install (Non-Interactive)

For a fast, fully automated install on Apple Silicon:

uvx voice-mode-install --yes
voicemode service install whisper
voicemode service install kokoro

What Gets Installed

Component	Size	Purpose
FFmpeg	~50MB	Audio processing (via Homebrew)
VoiceMode CLI	~10MB	Command-line tools
Whisper (base)	~150MB	Speech-to-text
Kokoro	~350MB	Text-to-speech

Implementation

1. Check architecture: uname -m (arm64 = Apple Silicon, recommended for local services)

2. Check what's already installed:

   which voicemode  # VoiceMode CLI
   which ffmpeg     # Audio processing
   

3. Install missing components:

   # Full install (installs ffmpeg, voicemode, and checks dependencies)
   uvx voice-mode-install --yes
   
   # Install local services
   voicemode service install whisper
   voicemode service install kokoro
   

4. Verify services are running:

   voicemode service status whisper
   voicemode service status kokoro
   

5. Reconnect MCP server: After installation, the VoiceMode MCP server needs to reconnect:

   • Run /mcp and select voicemode, then click "Reconnect", OR
   • Restart Claude Code

Whisper Model Selection

For Apple Silicon Macs with 16GB+ RAM, the large-v2 model is recommended:

Model	Download	RAM Usage	Accuracy
base	~150MB	~300MB	Good (default)
small	~460MB	~1GB	Better
large-v2	~3GB	~5GB	Best (recommended for 16GB+ RAM)
large-v3-turbo	~1.5GB	~3GB	Fast & accurate

To install the recommended model:

voicemode whisper install --model large-v2

Prerequisites

This install process assumes:

• UV - Python package manager (install: curl -LsSf https://astral.sh/uv/install.sh | sh)
• Homebrew - macOS package manager (install: brew.sh)

The VoiceMode installer will install Homebrew if missing on macOS.

For complete documentation, load the voicemode skill.