Local voice-to-text for your Mac.
Press a shortcut, speak, and the transcribed text appears wherever your cursor is – no cloud, no API keys, no subscriptions.
Version 0.4.0 · macOS 14+ · Apple Silicon & Intel
Also available for iOS
- Press Option+Space (or your custom shortcut)
- Speak – a floating preview shows your words in real time
- Press the shortcut again (or click the menu bar icon) to stop
- Your words are transcribed locally and pasted into the active app
Everything runs on-device using WhisperKit – no audio leaves your Mac.
- Real-time preview – A floating HUD shows confirmed and tentative text as you speak
- Local transcription – Powered by OpenAI's Whisper model running on Apple Neural Engine via WhisperKit
- Works in any app – Text is pasted wherever your cursor is (TextEdit, Notes, Slack, browser, etc.)
- Push-to-talk mode – Hold the shortcut to record, release to transcribe (configurable alongside toggle mode)
- Language auto-detection – Let Whisper identify the spoken language, or choose from 99 supported languages
- Smart filler removal – Strips "um", "uh", stutters, and context-aware fillers like "basically" and "sort of" without breaking real words
- Audio file import – Transcribe existing audio files (MP3, WAV, AIFF, OGG, FLAC) via the menu bar
- Auto-punctuation – Whisper adds punctuation naturally
- Transcription history – Review, search, and export past transcriptions
- Configurable shortcut – Change the hotkey in Settings
- Model selection – Choose between base, small, medium, or large Whisper models
- Launch at login – Optional, runs quietly in the menu bar
- Sound feedback – Audio cues when recording starts and stops
- Silence auto-stop – Configurable timeout (30s to 10 min, or never)
- Auto-mute playback – System audio is silenced while recording so YouTube, music, etc. don't interfere
- Trailing word capture – A 1-second buffer after you stop ensures your last words aren't clipped
- Download
Wave.dmgfrom the latest release - Open the DMG and drag Wave to Applications
- Launch from Applications
- Grant Microphone and Accessibility permissions when prompted
- The Whisper model (~460MB) downloads automatically on first launch
| Permission | Why |
|---|---|
| Microphone | Records your voice for transcription |
| Accessibility | Pastes transcribed text into the active app via simulated Cmd+V |
- Swift + AppKit (native macOS menu bar app)
- WhisperKit – Core ML-optimized Whisper inference
- KeyboardShortcuts – Global hotkey management
- No sandbox (required for Accessibility API + global hotkeys)
Found a bug or have a feature idea? Open an issue.
Made by santiagoalonso.com