Home

Ambient Listening

Speak the encounter. It's transcribed live and chunked to the backend for more accurate results.

Live captions need Chrome or Edge. Recording and backend transcription will still work — you just won't see live text while speaking.

Chunk settings

A chunk is cut once it has run for at least the minimum length and a silence gap is detected. If no gap ever appears, a chunk is force-cut 30s past the minimum so it never runs away.

Tap to start recording

Live captions have stopped responding — this is a known ~60s session limit on the browser's free speech service, not a data-loss issue. Recording and backend transcription continue normally; captions will resume automatically if the browser's throttle lifts.

Transcript

Waiting for chunks to finish processing...

cogneus.ai