Hey Product Hunt 👋 I'm Sergey, the maker of Apograph. Apograph runs Whisper models on the iPhone's GPU, so interviews, meetings and voice notes never leave the device. There's no account and no analytics SDK, and it works in Airplane Mode. A few things I cared about getting right: Transcripts you can fix. Tap a sentence to hear it, edit it in place, or retranscribe just that sentence with the Best model and compare before replacing anything. A spell checker that learns your names and terms and feeds them back to the model. Recordings are plain folders in the Files app (audio plus transcript.txt), not locked inside a database. Auto-stop ends a long recording at the minute you set, or before the phone runs out of space. Free covers everyday use: unlimited recordings, 99 languages, the Fast and Balanced models, text, Markdown and PDF export, and the Best model on five recordings so you can judge it yourself. Pro is a one-time purchase that adds unlimited Best, speaker separation and SRT/WebVTT subtitles. No subscription. I'd love to hear how it handles your language and accent, especially the ones the iPhone's built-in transcription doesn't support. And what's missing for your workflow? I'll be here all day answering.

Apograph

Private speech-to-text that never leaves your iPhone

Private speech-to-text that never leaves your iPhone