Discreet dictation for AI
Dictate to AI in public without broadcasting every word.
Talking to a chatbot in an open office is awkward. Voicecloak masks your voice while you speak and transcribes on-device straight to your clipboard — no new mic, no wearable.
Your voice, blurred by masking sound
A clearer recording for transcription
rec 00:07What you receive · clipboard
"refactor the auth middleware so the session check runs before rate limiting"
How the masking works
Voicecloak plays a layer of sound designed to make your words harder to make out nearby. Your iPhone then reduces that sound from the recording and transcribes your voice on-device — no special microphone or wearable required. Nothing is uploaded at any stage.
How it works
Speak at a whisper
Hold the button and talk low. Voicecloak plays masking sound in real time, so your words are harder to make out by anyone nearby.
Transcribed on-device
Speech recognition runs locally. No audio leaves the phone, no server sees the transcript, works on a plane.
Pasted where you need it
The text lands on your clipboard the moment you release. Paste into ChatGPT, Claude, Cursor, Slack — anything.
- Hardware needed
- None — the phone in your pocket
- Audio uploaded
- Zero bytes
- Platform
- iOS, via TestFlight
Want in? I'm looking for 25 TestFlight testers.
Tell me where you usually dictate — café, open office, shared apartment, hospital ward — and I'll send an invite. Reply by email or DM, whichever you prefer.