Audio is processed securely
When you dictate, your audio is sent over an encrypted connection to the ASR service purely to produce your transcript.
Private by default
Dhivehi ASR should feel safe before it feels clever. Here’s exactly how your audio and text are handled, in plain language.
Dictations aren’t stored by default
Once your Thaana text is returned, the audio is not kept. There’s no silent recording archive sitting behind the app.
History is opt-in, later
If we add saved history in future, it will be something you turn on — and you’ll be able to delete it whenever you want.
No analytics on your words
We never use your transcript text for analytics without consent. Product events record actions like ‘recording started’, never the content you dictate.
Retention & deletion
Default retention is none: nothing to delete because nothing is stored. Team/Institution customers get explicit retention and deletion controls.
You stay in control
Personal dictionary words and saved snippets are settings you own on your device or account — yours to edit or remove.
Questions from schools, offices & government
Is my audio stored?
No. Audio is processed securely to produce your transcript and is not stored by default. History is something you would opt in to later, not something that happens automatically.
Is my transcript used for analytics?
We never use your transcript text for analytics without consent. Product analytics only record named events like ‘recording started’ — never the words you dictate.
How long is anything retained?
By default nothing is retained after your transcript is returned. If you later opt in to saved history, you stay in control and can delete it at any time.
Can schools, offices, and government use it?
Yes. The Team / Institution plan adds shared vocabulary, admin controls, and privacy controls designed for education, government, and media workflows.
Which platforms are supported?
There’s a desktop app for macOS and Windows and a web demo today. Android and iOS keyboards are on the roadmap — join the mobile waitlist to hear first.
Do I need to switch keyboards?
No. You speak in Dhivehi and get clean Thaana text you can paste anywhere, without flipping between Latin and Thaana layouts.