https://github.com/elifuzz/vox
https://github.com/elifuzz/vox
Last synced: about 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/elifuzz/vox
- Owner: EliFuzz
- License: apache-2.0
- Created: 2026-03-29T01:13:01.000Z (3 months ago)
- Default Branch: main
- Last Pushed: 2026-04-03T22:23:46.000Z (2 months ago)
- Last Synced: 2026-04-04T00:26:10.048Z (2 months ago)
- Language: Rust
- Size: 132 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Vox
**Vox** is a small macOS menu bar app for local speech-to-text.
Hold a shortcut, talk, and it drops the text right where your cursor already is.
No giant window. No extra editor. No copy/paste dance.
> Speak. Release. Keep moving.
## Features
- **local transcription** - your speech becomes text on-device
- **cursor-first output** - words land in the app you're already using
- **two recording styles** - push-to-talk or toggle-to-talk (Press `Fn + Left Control` to talk, release to paste. Or double-press to toggle)
- **pick your mic, pick your model** - without digging through settings files
## Quick start
1. Use GitHub Releases to grab the latest version of **Vox** for macOS. Just download, unzip, and move it to your Applications folder.
2. Open **Vox**
3. Allow **Accessibility** on macOS _(and **Input Monitoring** too, if macOS asks)_
4. Hold **Fn + Left Control** to talk
5. Release to paste text or wait a half-second for it to paste automatically while you keep talking
6. Double-press the shortcut for **toggle-to-talk**
## Under the hood
| area | details |
| ---------------- | -------------------------------------------- |
| input | selectable microphone device |
| speech detection | VAD-gated recording |
| transcription | local ASR pipeline |
| models | Multilingual, English, English Small |
| engines | Parakeet, Moonshine |
| post-processing | sentence normalization + punctuation cleanup |
| output | paste at cursor + restore previous clipboard |
| settings | `~/.vox/settings.json` |
| model cache | `~/.vox/models/` |
| model delivery | downloads models when needed |
## project layout
- `src/main.rs` - boot, workers, update checks
- `src/processing.rs` - speech segmentation + transcription flow
- `src/tray/` - menu bar state and menus
- `src/platform/macos/` - keyboard hooks, Accessibility, paste, CoreML bridge
- `src/model_download.rs` - model fetch + unzip
- `src/settings.rs` - mode, shortcut, device, model