how to use speaker diarization with deepgram stt

Thanks for reaching out, Usama. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

which STTs support diarization

Hey there! Just so you know, in LiveKit, speaker diarization is currently supported only by Speechmatics STT. If you want to give it a try, you can turn it on by setting

diarization="speaker"

in your STT configuration. You can also tweak things a bit with optional settings like

max_speakers

and

speaker_sensitivity

to help fine-tune how speakers are detected. Hope that helps! Sources: Speech-to-text (STT) integrations | LiveKit Docs | Speechmatics STT integration guide | LiveKit Docs | pipeline-stt/diarization.py