Yo boyz vapi is quite good and nice tool that i didn’t expect to be, now im using in this way
Deepgram as a transcriper
Gpt for text generation
11labs for text to speech
One interesting thing i found out is the transcriper gives u a word by word timing however speech to text service aka 11 labs gives only for the whole sentence rather than words
I checked 11 labs docs and they offer the word by word timing in their apis i wonder how can i get it in vapi?
That’s currently how it looks like
"role": "bot",
"message": "Or",
"time": 1759011498858,
"endTime": 1759011499098,
"secondsFromStart": 13.003,
"duration": 240,
"source": ""