Data and permissions

Microphone

Required only when you start a recording. The app keeps a visible recording state and does not support hidden recording.

Local storage

Used for encrypted recordings, transcripts, minutes, downloaded speech models, settings, and user-created templates. Temporary decrypted files may be created for playback, transcription, sharing, or export and are intended to be removed after the operation.

Network

Not required for an already-installed model’s local transcription and minutes generation. Used for user-initiated model downloads, optional Google Drive, platform purchases, and optional consented diagnostics.

Google account and Drive

Optional. The app requests access only when you choose to connect Google and use the selected Drive features. Nothing is uploaded by default.

OpenAI

The app does not use the OpenAI API and does not send recordings, transcripts, or minutes to OpenAI. Whisper model weights execute locally through sherpa-onnx.

Last updated: 28 August 2026