vtype 0.1.2

Offline voice input — speak into any focused window, no cloud required

vtype

Offline voice input for any focused window. Press Ctrl+Alt+V (Windows/Linux) or Control+Option+V (Mac), speak, and your words are typed instantly — no cloud, no API key, no subscription.

Works in Claude Code terminal, VS Code, any text editor, browser, chat app — whatever window is focused.

---

Install

pip install vtype

First run downloads the Whisper speech model (~150 MB). Subsequent runs are instant.

Usage

vtype

• Press Ctrl+Alt+V (Windows/Linux) or Control+Option+V (Mac) — starts listening
• Speak naturally
• Silence for ~1 second → transcribes and types into the focused window
• Press the hotkey again to cancel
• Ctrl+C to quit

Requirements

• Python 3.9+
• A microphone
• FFmpeg (for audio processing)

FFmpeg install:

Platform	Command
Windows	winget install ffmpeg
macOS	brew install ffmpeg
Linux	sudo apt install ffmpeg

How it works

• Speech detection runs locally using RMS-based VAD (no network calls)
• Transcription uses OpenAI Whisper (base model, runs on CPU)
• Text is typed via OS-level keyboard simulation — works in any window including terminals and browser tabs

Privacy

Everything runs on your machine. Audio never leaves your device.

License

MIT