Server turn detection for a transcription session: OpenAI's voice activity detector decides where each utterance ends and commits it.
Optional
Activation threshold in [0, 1]; a higher value needs louder speech.
0.5 Copy
0.5
Audio kept before detected speech, in milliseconds.
600 Copy
600
Silence that ends an utterance, in milliseconds.
350 Copy
350
Server turn detection for a transcription session: OpenAI's voice activity detector decides where each utterance ends and commits it.