Mastra doesn't support speaker diarization despite the voice AI supporting it.
I was using Deepgram which supports speaker diarization. I have tried a few ways to enable the diarization option. But it's not working. I have tried adding
{ diarize: true }
{ diarize: true }
with
.listen()
.listen()
and
listeningModel
listeningModel
. It's not working. Deepgram does transcribe the audio. But it doesn't provide any diarization.