Trust

Security

Dexlyy Voice is engineered so the most private option is also the default. Here's how it works.

On-device by default

Free mode runs a local speech-to-text model on your own machine. Audio is transcribed offline and never transmitted — no servers, no uploads, no exceptions.

Opt-in cloud, encrypted in transit

Cloud mode is something you turn on deliberately. When enabled, audio is sent over TLS-encrypted connections only to perform your transcription or translation, then discarded. It is never used to train models.

Hardened local sidecar

The app's local processing service binds to 127.0.0.1 only — it is never exposed to your network. Production builds run with a per-session security token and strip privileged service-role keys, so nothing on your device can act with elevated server permissions.

Your data, your control

Responsible disclosure

Found a vulnerability? We want to hear from you. Email security@dexlyy.com with details and we'll respond promptly. Please give us reasonable time to remediate before public disclosure.