It would be useful to support more Deepgram transcription features. Being able to toggle diarization would be most useful to me, but also various formatting options the API offers: utterances, punctuation, paragraphs, etc.
Currently if I use the hyperaudio editor with Deepgram I get diarization if I use the 'Enhanced' model, but not the 'Nova-2' or Whisper options.