Most voice-to-text tools assume that uploading audio to the cloud is acceptable. That is convenient, but it is not always the right default for meetings, interviews, lectures, client notes, or personal journals.
I have been working on VoiceScriber, an iPhone app that records and transcribes directly on-device. The goal is simple: turn audio into searchable notes without requiring an internet connection or sending the original recording to a server.
The app currently supports:
- Offline transcription on iPhone
- Recording and transcription without Wi-Fi or cellular data
- On-device storage for recordings, transcripts, and notes
- Search, tags, and export for working with the transcript later
- More than 100 transcription languages
- Longer recording sessions for meetings and lectures
The main product decision was to treat privacy as the default architecture instead of a settings toggle. If the user records something sensitive, the safest path should be the normal path.
There are tradeoffs. On-device transcription depends on the phone's capabilities, and offline apps need to be careful with storage, long recordings, interruption handling, and battery use. But for people who care about privacy or need transcription when the network is unreliable, those tradeoffs are worth exploring.
The app is here: https://voicescriber.com/
I would be interested in feedback from iOS developers and privacy-minded users:
- Is the offline/on-device positioning clear?
- What would you expect from a private transcription workflow?
- Which export or note organization features matter most after the transcript is created?
Top comments (0)