Effective date: October 13, 2026
TalkText ("we", "us") turns your recordings into text. It is built so that your audio and transcripts stay on your device. This policy explains what that means.
Your recordings and transcripts
- Transcription happens on your iPhone. Speech recognition and speaker detection run on your device. We do not upload your audio, video, transcripts or summaries to our servers, and we do not operate servers that receive them.
- Storage. Recordings, imported files, transcripts and summaries are stored in the app's private storage on your device. They may be included in your own iCloud or computer backups if you have device backups turned on; those backups are controlled by Apple and you, not by us.
- iCloud sync (optional, off by default). If you turn on "Sync with iCloud" in Settings, your recordings and transcripts are stored in your own private iCloud database (Apple CloudKit) so they appear on your other devices. This data is held by Apple under your Apple Account; we cannot access it.
- Summaries. Optional summaries use Apple Intelligence on your device. See Apple's privacy policy for how Apple handles on-device model features.
- Sharing. Text, subtitle (SRT) or audio files leave the app only when you choose to share or export them. When you share a file into TalkText from another app, it is copied into the app's storage on your device.
- Widget. The home screen widget shows only counts and the title of your latest recording, read from your device.
- Deleting. Deleting a recording in the app removes its audio and transcript from the device. You can also have audio deleted automatically after transcription in Settings.
Data we do not collect
We do not require an account and do not collect your name, email, contacts, location, or the content of your recordings. The app contains no advertising and no third-party analytics or tracking.
Speech models
The speech-recognition and speaker-detection models are included in the app. In rare cases where a model is missing or needs to be replaced, the app downloads it from Hugging Face; that download sends a standard network request (such as your IP address and device type) to Hugging Face, as with any download. No audio or text is sent.
Purchases
Subscriptions and purchases are processed by Apple. We use RevenueCat to check whether your purchases are active. RevenueCat receives an anonymous app user ID and your purchase history from Apple; it is not linked to your identity. See RevenueCat's privacy policy at https://www.revenuecat.com/privacy.
Permissions
- Microphone: to record audio you choose to transcribe. Used only while you are recording.
- Files: to import audio or video you select.
- Speech recognition (live preview): the live text shown while recording uses Apple's on-device speech framework. Audio is not sent to Apple or to us for this.