- What Dictiq collects
- Nothing. There is no backend server, no user database, no account and no login.
- Where processing happens
- On your machine by default. If you switch on your own endpoint, the stage you point at it runs wherever that endpoint is. Dictiq never captures, reads or transmits screen content. It may check only which window has focus, by process name or title, to change its own behaviour for that app, and that check happens on your machine.
- What leaves your machine
- Nothing you dictate, unless you switch on your own endpoint. That is off by default and part of the one-time unlock: you supply the address and the key, and Dictiq then sends your text there and nowhere else, and keeps no copy of it. The only other network use is downloading the language models you choose and Microsoft Store licence validation. The settings window is drawn with the Edge WebView2 component built into Windows. That component talks to services of its own, including diagnostics it sends Microsoft under your Windows Diagnostics & feedback setting, and it cannot be told not to. It is Windows', not ours: it never sees your dictation, and nothing from it reaches us.
- Effective date
- 2026-07-10(provisional while the product is in development; confirmed at its Store release)