Skip to main content

Live Captions

Aria provides real-time captions during any room or session using on-device speech recognition.

How it works​

Live captions use on-device speech recognition on supported devices. Audio used for captioning is processed locally and is not uploaded for caption generation.

Enabling captions​

  1. While in a room or session, tap the captions toggle
  2. Captions appear at the bottom of the screen as participants speak
  3. Tap again to turn them off

Privacy​

  • Speech recognition for captions runs on-device
  • Audio is not sent to Aria servers for caption generation
  • Captions are not saved as part of room or session history

Availability​

Live captions are available in all four space types: Private Rooms, Public Campfires, Public Sessions, and Subscriber Sessions.

Accuracy​

Caption accuracy depends on audio quality, background noise, accent, and speaking speed. Captions work best with clear audio in quiet environments.

Captions and AI Voice Agents​

Live captions also work when AI voice agents are speaking in a session. Agent speech is captioned just like human speech, making agent conversations accessible to all participants. Learn more about AI voice agents.