I wanted to know other people s solution

[rare narwhal]

I have wanted a separate channel for API vs local model. I am looking into creating a streaming API for Whisper to run in real time since I'm not aware of any other option. I am curious about Deepgram, @ashen vale . I assume that costs money? Can they use a Whisper model that I fine-tune on my own data? (I get much higher accuracy by doing fine-tuning.)

[ashen vale]

DG costs money indeed, I'm sure there may be local streaming solutions but I'm not too familiar with that unfortunately

[glad grotto]

There is no channel for the local model? Sad