Privacy

Audio handling, in plain English.

Audio is sent to typwrtr Cloud every time you dictate. We forward it to our transcription provider, return the text, and discard the audio. We never store recordings. We never use your audio to train models.

What we keep

• Your Google account email + identifier (sub claim) — to look up your subscription.
• Per-request: request id, clip duration in seconds, an account hash, HTTP status. No transcripts, no audio.
• Monthly usage minutes, for billing visibility.

What we never keep

• The audio bytes.
• The transcript text.
• Any signal that ties a transcript to your account beyond the request log above.

Local data

The desktop app stores transcripts and the self-learning corpus on your machine, in your user profile. Wipe from Learning → Memory inside the app.

Payment data

Payments are processed by Dodo Payments as Merchant of Record. We never see your card or UPI details. Tax handling, refunds, and disputes live in your customer portal.