Is input_audio_transcription working now?

[meebox]

I update the session with

               "input_audio_transcription": {
                   "model": "gpt-4o-transcribe"
               }

but can’t get transcribe.