Voice to text.
Nothing leaves your Mac.
State-of-the-art AI transcription running entirely on your computer. Local means private. Local means fast. No cloud. No account. No data collection. Free to download.
Transcribing...
The quick brown fox jumps over the
How it works
Hold Fn
Press and hold the Fn key to start recording. A green dot appears in your menu bar.
Speak naturally
Talk at your normal pace. You'll see a live streaming preview of the transcription as you speak.
Text appears at cursor
Release the key and the final transcription is inserted right where your cursor is. Any app, any field.
Everything you need, nothing you don't
Local & Private
Your audio goes from microphone to Apple Silicon to text. The network is never involved. No account, no cloud, no tracking.
Streaming Preview
See your words appear in real time as you speak. The transcription streams directly into your active text field.
Smart Vocabulary
Add custom word corrections for names, jargon, or terms Whisper gets wrong. Corrections apply automatically.
Voice Commands
Say "WhisperFree correct X to Y" while recording to add vocabulary corrections hands-free.
Works Everywhere
Text is inserted at your cursor in any app — email, Slack, Notes, VS Code, your browser. If you can type there, WhisperFree works there.
Fast by Design
No upload, no round-trip, no waiting on a server. Local processing means transcription starts in milliseconds, not seconds.
Your voice stays on your Mac
There is no server. WhisperFree processes everything on your Mac using Apple Silicon. This isn't a privacy policy — it's the architecture.
- AI model runs on your Mac
- Audio processed in memory only
- No network requests during transcription
- No account required
- No telemetry or analytics
- Audio sent to remote servers
- Data stored and potentially used for training
- Requires internet connection
- Requires account and often payment
- Usage tracked and logged
Private because there's nothing to compromise. Fast because there's no server in the way.
System requirements
Operating system
macOS 14.0+
(Sonoma or later)
Processor
Apple Silicon required
(M1, M2, M3, M4, or later)
Disk space
~150 MB
for Whisper model
Permissions needed: Microphone access (for recording) and Accessibility access (for inserting text at your cursor). A guided setup walks you through granting both on first launch.