Notate: Speech to text that never leaves your device

Whisper transcription that runs entirely on your iPhone, Apple Watch or Mac. The models ship inside the app, so it works offline from the first launch. The word being heard lights up as you talk, and it keeps everything said. On-device summaries with action items, search across every recording, export as text, Markdown or SRT. No account, no cloud, no subscription. $0.99 once.

Hi, I'm Josh, I built Voxprint. I wanted dictation on my phone that I could trust with anything. Every app I tried sent the audio somewhere, asked for an account, or wanted eight dollars a month for what is really a model running on hardware I already own. So I put Whisper on the device itself. Voxprint runs WhisperKit locally on iPhone and Mac. Audio is decoded in two second windows as you speak, so words show up while you are still talking rather than after you stop. It picks a model that fits your device's memory, caches it somewhere iOS will not purge, and works on a plane or in a basement because there is no network call to make. Nothing you say is uploaded, ever. There is no sign-up and no tracking. It handles more than voice. Drop in a podcast, a lecture, or a song and it writes out the words, with auto-detect across 12 languages. Live microphone transcription is free with no limits. The one paid thing is Voxprint Pro, a single one dollar purchase that unlocks unlimited audio file transcription and the Small model, the most accurate one. You buy it once. I'd rather sell a tool than rent one. The web app and the whole codebase are open. Happy to answer anything about the on-device pipeline.

Update: Voxprint is now Notate, and 2.0 is out. The models ship inside the app now, so it works offline from the first launch with no download. New in 2.0: the word being heard lights up as you talk, on-device summaries with action items, search across every recording, export as text, Markdown or SRT, Shortcuts, and an Apple Watch app. No Pro tier anymore, just $0.99 once.