# Apograph

> Apograph records, transcribes and labels who said what — entirely on your iPhone, in 99 languages, with nothing uploaded.

Apograph records, transcribes, tells apart who said what — and gives you a transcript you can edit in place: fix a word, a name, a whole sentence. All on the device, in 99 languages; everything works in Airplane Mode. Nothing is uploaded, ever. Pro is a one-time purchase (Everything in Free; Best model — the highest transcription accuracy; Unlimited speaker separation — who said what; SRT and WebVTT subtitle export; Unlimited Best re-transcriptions). Requires iOS 18.6 or later. Publisher: Sinergy Labs. Support: support@getapograph.io.

## Pages (English)

- [Landing page](https://getapograph.io/en/): Private, on-device transcription for iPhone
- [Privacy policy](https://getapograph.io/en/privacy/): Apograph is built so that your recordings, transcripts and everything derived from them stay on your iPhone. The app has no account system, no server of its own that receives user content, and no analytics or advertising code. This policy explains exactly what stays on the device, the two things that do use the network, and what Apple handles on our behalf.
- [Support](https://getapograph.io/en/support/): Apograph is made by a small team. Write to us and a person will answer.
- [Full text of the English pages](https://getapograph.io/llms-full.txt)

## Languages

Every page exists in each of these languages at https://getapograph.io/<code>/…; the English privacy policy is the authoritative one.

- English (en): https://getapograph.io/en/
- Deutsch (de): https://getapograph.io/de/
- Español (es): https://getapograph.io/es/
- Français (fr): https://getapograph.io/fr/
- Italiano (it): https://getapograph.io/it/
- Português (pt-BR): https://getapograph.io/pt/
- Русский (ru): https://getapograph.io/ru/
- 日本語 (ja): https://getapograph.io/ja/
- 한국어 (ko): https://getapograph.io/ko/
- 简体中文 (zh-Hans): https://getapograph.io/zh/
- العربية (ar): https://getapograph.io/ar/

## Facts

- On-device transcription with OpenAI Whisper (whisper.cpp) in 99 languages; speaker separation with sherpa-onnx; nothing is uploaded.
- Free: unlimited recordings and imports, Fast and Balanced models, text/Markdown/PDF export. Pro (one-time purchase): Best model, speaker separation, SRT/WebVTT subtitles.
- The only network traffic is fetching the model catalog, downloading a chosen model, and the App Store handling a purchase.