Official documentation
No Type: data and speech processing
No Type, 2026
Read the originalRead on October 20, 2026
- What it is
- The No Type page on data and speech processing. It says where speech is recognized and where text is processed, what is stored on the device, and what the app's usage statistics send to the maker.
- How much weight it carries
- The maker's own privacy statement. It describes the app's design; the policies of any cloud provider the user connects govern the data sent to that provider.
Facts taken from this source
- No Type says audio is deleted after successful recognition.
“Audio is deleted after successful recognition. Only the ten most recent recordings with failed recognition and no text are kept for retrying.” (Section 'History and settings')
- No Type's privacy page says recordings, history and settings are stored on the user's device.
“History, recordings, dictionary entries, snippets, settings and downloaded models are stored on your device.” (Section 'History and settings')
- No Type's privacy page says that to rewrite text, the user chooses a local model in the app or turns that stage off.
“To rewrite text, choose a local model in the app or turn that stage off.” (Section 'Local processing')
- No Type's privacy page says Whisper, Parakeet and GigaAM recognize audio on the user's device on macOS and Windows.
“Whisper, Parakeet and GigaAM on macOS and Windows recognize audio on your device.” (Section 'Local processing')
- No Type says it keeps only the ten most recent recordings with failed recognition and no text, for retrying.
“Only the ten most recent recordings with failed recognition and no text are kept for retrying.” (Section 'History and settings')