No data collection by default
VoiceSift works fully offline and requires no account. We do not collect, transmit or sell any personal information. There are no analytics or advertising SDKs in the app.
Audio never leaves your device
All audio is recorded and stored locally. Speech recognition, speaker identification and keyword extraction run entirely on your device. Audio is never uploaded, under any feature or circumstance.
Encrypted local storage
Recordings, transcripts, keywords and speaker profiles are stored in an encrypted local database on your device only. Recordings can be exported to your Downloads folder at your request.
Optional AI Minutes
Minutes generation is disabled by default. If you choose to enable it, the transcript text (not audio) is sent over HTTPS to the large-model provider you have configured with your own API key, solely to generate the minutes. The app displays a clear notice before sending and specifically warns against submitting legally sensitive recordings. You can disable the feature at any time.
Third-party components
The app integrates a small number of open-source components (see the SDK List) that run locally and do not transmit personal data.
Contact
For privacy questions, contact us at grswxt@126.com.