Privacy Policy

Transcripts for iPhone and iPad Last updated: 15 August 2026

The short version

Transcripts collects nothing. It has no servers, no accounts, no analytics, and no advertising. It contains no networking code of any kind — not disabled, not optional: absent. Your recordings, transcripts, and settings stay on your device and in the folder you choose.

What the app stores, and where

• Audio recordings are written on your device as ordinary .m4a files.
• Transcripts are stored alongside them as plain text in a small .json file per recording.
• Settings (your chosen folder, app preferences) are stored in the app's own preferences on your device.

When you choose a destination folder, recordings are copied there. If that folder is inside a cloud service you already use — iCloud Drive, OneDrive, Dropbox — then that service syncs the files under its privacy policy and your account with them, exactly as it would for any other file you put in that folder. Transcripts is not party to that transfer; it writes a file to a folder your device has given it access to.

Speech recognition

Live transcription uses Apple's on-device speech recognition. Transcripts sets Apple's requiresOnDeviceRecognition flag, and additionally refuses to transcribe at all on a device that cannot recognize speech locally — in that case it turns live text off and tells you why, rather than falling back to server-based recognition. Your audio is never sent to Apple's speech servers, or anyone else's.

On devices with Apple Intelligence, summaries are generated by Apple's on-device foundation model. This also runs locally.

Microphone

Transcripts uses the microphone only while you are recording, and only after you grant permission through iOS. You can revoke it at any time in Settings ▸ Privacy & Security ▸ Microphone. Recording continues in the background so a capture survives the screen locking — this is the point of a pocket recorder, and iOS shows a persistent indicator whenever it is happening.

What is never collected

No personal information, no contacts, no location, no device identifiers, no usage analytics, no crash reporting, no advertising identifiers, and no tracking of any kind. Transcripts does not use third-party SDKs — it has no third-party dependencies at all.

This website

The apps collect nothing — that is a property of how they are built. The iOS app contains no networking code at all, and the Mac app's only outbound request is checking whether a newer version exists.

This website counts visits using Cloudflare Web Analytics, which measures page views, referrers and rough location without cookies and without building a profile of you. There is no advertising network involved, nothing is shared with Google, and there is nothing here to opt out of because nothing follows you anywhere.

Children

Transcripts is not directed at children and collects no data from anyone, including children.

Credits

Speaker attribution on the Mac uses on-device diarization models from FluidInference, derived from pyannote, both under CC BY 4.0. They are downloaded once, run entirely on your Mac, and are used unmodified.

Your recordings are your responsibility

Recording laws vary by country and by state. Some jurisdictions require the consent of every participant before a conversation may be recorded. Transcripts gives you a recorder; deciding whether you may lawfully record a given conversation, and obtaining any consent required, is yours.

Changes

If this policy changes, the updated version will be posted here with a new date. Because the app collects nothing, changes are expected to be rare.

Contact

Questions about this policy: [email protected]

Everything above is a description of how the app is built, not a promise about how it is operated — there is no server to operate. If any of it ever stops being true, it will stop being true here first.