We’ve been using gpt-4o-transcribe a lot. Thousands of hours transcribed and today we switched to gpt-4o-transcribe-diarize and it feels soo slow.
At first I thought that it was just my impression while testing this model in dev environment, but once we release it in prod, I instantly noticed how audio files that are in process of being transcribed started piling up.
Has anyone else notice that gpt-4o-transcribe-diarize is way slower?