• Audio 엔드포인트 → .speech.create()

response = client.audio.speech.create(

    model="gpt-4o-mini-tts",

    voice="onyx",

    input="Creating human-like speech is now possible with just a few lines of code.
    Pretty neat, right?"

)


response.stream_to_file("output.mp3")

• response_format: "mp3", "opus", "aac", "flac", "wav", "pcm"