VoiceScribe is an offline voice recorder that transcribes and summarizes audio entirely on your device. It was built so that your recordings never have to leave your phone — and they don't.
Nothing. VoiceScribe has no user accounts, no sign-in, no analytics, no advertising SDKs, and no trackers. We do not collect, transmit, sell, or share any personal data.
Audio you record or import, and the transcripts, summaries, and speaker labels generated from it, are processed by AI models running locally on your phone and are stored only in VoiceScribe's private app storage on your device. They are never uploaded anywhere. They are removed when you delete a recording in the app or uninstall the app.
VoiceScribe requests microphone access solely to record audio when you press the record button. Recording is always user-initiated, shown with a visible notification while active, and never happens in the background without you starting it.
VoiceScribe works fully offline, including on first launch: the default transcription model is included in the app itself, so nothing needs to be downloaded to use it. The network is used for exactly one optional thing:
Transcripts, summaries, and exported PDF/Word documents leave the app only when you explicitly share them through the Android share sheet, to an app you choose. VoiceScribe does not transmit them on its own.
VoiceScribe is free and open-source software (MIT licensed). There are no in-app purchases, no subscriptions, and no advertising. You can read the complete source code — and verify every claim on this page — at github.com/llassan/voice_scribe.
VoiceScribe is not directed at children under 13 and collects no data from anyone, including children.
If this policy changes, the updated version will be posted at this URL with a new effective date. Because the app collects no data, changes are expected to be rare.
Questions about this policy: [email protected]