I ultimately need to convert the word into visemes that play on an Avatar. For this I just need the word, and time stamp start/end per word

Following up on this

Thanks - is it in the development roadmap?

Not at this time

@User would it be possible to add a parameter to the transcriber parameters on the Create Assistant endpoint so that VAPI could return a timestamped transcript for the call? As it stands right now, I'm having to get the recordingUrl from the call and then send an additional deepgram call to get another transcript with timestamps. "transcriber": { "provider": "deepgram", "model": "nova-2", "language": "bg", "smartFormat": true, "keywords": [ "" ] },

as of now not possible to add param to the transcriber,