LAST UPDATED 31 JULY 2026

Privacy, plainly.

VoiceCrisp is designed to process dictation on your iPhone, rather than sending it to a service operated by us.

What VoiceCrisp processes

VoiceCrisp accesses your microphone only when you choose to record. Your speech is transcribed and corrected for recognition mistakes, punctuation and text coherence on your device. It does not summarise your dictation or change its meaning. Audio is deleted after transcription.

What we collect

We do not operate user accounts, do not use advertising or analytics SDKs, and do not collect your recordings, transcripts, contacts, usage data or identifiers.

Network use

The app may download a speech-recognition model when you choose one. Once a model is downloaded, transcription and correction work locally. Network information required to perform that download is handled by the download connection and is not retained by us.

Keyboard extension

The VoiceCrisp keyboard does not record audio. It exists solely to insert dictations you have already created in VoiceCrisp into a text field you choose.

Files and sharing

If you enable Markdown transcript files or choose to share text, those actions are controlled by you and use the destinations you select. VoiceCrisp does not receive a copy.

Changes

We will update this page if the app's privacy practices change. Material changes will be reflected in the App Store privacy information before release.

Contact

Questions: marcos@devrelbridge.com.