OpenAIRealtime
OpenAI Realtime speech-to-speech (S2S) model.Constructor
cleanup
close
getRuntimeConfig
SUPPORTED_VOICES
Voice ids supported by the OpenAI Realtime model.TurnDetectionConfig
Server-side voice-activity turn-detection settings. Passnull to disable.
Fields
Detection mode. Default
'server_vad'.Speech-probability threshold in [0, 1]. Default
0.5.Audio padding (ms) prepended before detected speech. Default
300.Trailing silence (ms) that ends a turn. Default
200.Auto-create a model response when a turn ends. Default
true.Allow user speech to interrupt the model’s response. Default
true.InputAudioTranscriptionConfig
Input-audio transcription settings. Passnull to disable transcription.
Fields
Transcription model id. Default
'gpt-4o-mini-transcribe'.OpenAIRealtimeConfig
Configuration for OpenAIRealtime.Fields
Realtime model id. Default
'gpt-4o-realtime-preview'.Voice id (see SUPPORTED_VOICES). Default
'alloy'.Response modalities. Default
['text', 'audio'].Sampling temperature. Default
0.8.Max output tokens per response. Default the string
'inf' (unlimited);
null omits the param.Turn-detection settings, or
null to disable. Defaults applied when omitted.Input-audio transcription settings, or
null to disable. Default model applied when omitted.Tool-selection strategy. Default
'auto'.