Multi-lingual, work-safe transcription.
Other tools upload audio. Vox doesn’t.
| Vox | Wispr Flow | Granola | Apple Dictation | |
|---|---|---|---|---|
| Price | Free (beta, limited time) | $15/month | $14/month | Free |
| Local / private | ✓ | ✗ | ✗ | ~ |
| Any text field | ✓ | ~ | ✗ | ~ |
| Fourteen languages | ✓ | ✗ | ✗ | ~ |
| Works offline | ✓ | ✗ | ✗ | ~ |
Apple Silicon Macs running macOS 11 or newer. Transcription speeds are comparable to Wispr Flow on M2 chips and higher.
Yes. Vox is free to download during the public beta (limited time). There is no subscription.
The Vox installer (DMG) is about 118 MB. On first run, setup also downloads the on-device speech model (~1.5 GB) once.
No. Internet is needed for first-run model download. After that, transcription works offline.
Microphone (record while you hold the hotkey), Accessibility (type at your cursor in any app), and Input Monitoring only if you choose Fn as your hotkey.
No. Audio is transcribed on your Mac and is not uploaded.
On your Mac, locally, in two places:
The Settings → System → Uninstall button removes both, plus cached files and macOS permission grants, in one click.
Fourteen: English, French, German, Italian, Spanish, Portuguese, Greek, Dutch, Polish, Chinese (Mandarin), Japanese, Korean, Vietnamese, and Arabic.
Put the cursor in a text field, hold your hotkey, speak, and release. Vox types where your cursor is focused.