Privacy
Last updated July 25, 2026.
What we collect
DictatorFlow collects account details you provide, payment and subscription status from billing providers, API usage records, and technical logs needed to operate the service.
Audio and transcripts
Audio sent to the cloud API is processed to return a transcript. By default, it is not eligible for training publicly downloadable models. An account owner may separately opt in to contribute future cloud dictation for public model weights, and may revoke that choice for future collection. Legacy recordings and general product-improvement preferences are not treated as public-model consent. Revocation removes queued contribution files that have not entered an already published model version; published weights cannot be retroactively withdrawn. Desktop local transcription can run on your machine without sending audio to us.
How to reach us
For privacy questions, email [email protected].