Privacy
Local-first by design.
This page explains how Voice Inbox handles audio, documents, and network access in user-facing terms. It is not a substitute for legal review before app store publication.
Audio recognition
Voice Inbox runs speech recognition locally on device after the speech model is installed. On Android, audio files are read from the folder selected by the user. On iOS, the current MVP works with imported audio files stored in the app-local catalog.
Network access
Network access is needed for the first speech model download. The app should not be described as never using network access.
Selected files and folders
The app uses platform document access selected by the user. Android uses one output .txt or .md document and one audio folder.
iOS uses document picker access for imported audio and selected output documents in the current MVP.
Voice Inbox appends recognized text only to the selected output document.
Catalog storage
Voice Inbox keeps a private local catalog of discovered or imported audio files and processing state so it can distinguish New and Processed files.
External providers and platform services
Document providers, media decoders, storage behavior, and background scheduling are provided by the operating system or selected providers. Voice Inbox cannot control or make privacy claims on behalf of those external components.