openai.internal.CcMessageAudio

Audio output (`gpt-4o-audio-*`), requested through the `request_body` passthrough with `{"modalities":["text","audio"],"audio":{"voice":…,"format":…}}`. The response carries the spoken bytes and their transcript; the format is NOT echoed back, so the mime type is recovered from the request (see `_chat_audio_output_mime`).

Reference version

Signature

class openai.internal.CcMessageAudio

Audio output (gpt-4o-audio-*), requested through the request_body passthrough with {"modalities":["text","audio"],"audio":{"voice":…,"format":…}}. The response carries the spoken bytes and their transcript; the format is NOT echoed back, so the mime type is recovered from the request (see _chat_audio_output_mime).

Source:<builtin>/openai/ns_internal/chat.bamlbytes 15582–15689

Fields

id

string | null

data

string | null

transcript

string | null

expires_at

int | null