Your data and privacy¶
Recognition runs on your own phone or computer. Audio and transcripts aren't uploaded. The apps go online to download models and to check for new versions. The privacy page has the full statement.
On a computer¶
Settings → Your data lists every place below, with sizes, and has buttons to open them.
| Linux | Windows | |
|---|---|---|
| Settings | ~/.config/hebrish/config.json |
%APPDATA%\Hebrish\config.json |
| Dictionary | ~/.config/hebrish/dictionary.json |
%APPDATA%\Hebrish\dictionary.json |
| History, models | ~/.local/share/hebrish/ |
%LOCALAPPDATA%\Hebrish\ |
| Saved transcripts and recordings | Documents/Hebrish/ |
Documents\Hebrish\ |
History holds text only, not audio, unless you turn on Maintainer mode. Save audio on the Transcription tab keeps recordings in your Hebrish folder.
Back up and export¶
- Back up… (Settings → Your data, or the History tab) saves one archive of your history, saved transcripts, dictionary and settings. Models and recordings are left out.
- Export… on the History tab saves your history to a file.
Delete¶
- Clear history on the History tab deletes all transcripts.
- Each row in History has its own delete button.
- Turn off Keep a history of every transcript to stop keeping new ones.
If you turn on the Local API's agent control, programs on your computer can read your history. It's off by default.
On a phone¶
Models, settings and history stay in the app's own storage, unless you choose a shared model folder or a transcripts folder. Export CSV and Clear history are under Transcription history.
Help improve Hebrish¶
Nothing is sent to help train the models unless you choose to send it yourself. See Get involved.