Text to speech

Every native speech request uses client.audio.speech and the teen-v1 model.

speech = client.audio.speech.create(
    model="teen-v1",
    voice="default",
    input="Generate speech with Teen V1.",
    response_format="mp3",
)

Build your speech flow

1. Select default, a public voice ID, or inline reference audio.
2. Send text with audio.speech.create().
3. Save the complete response or consume it while it is generated.
4. Add generation controls only when the default output needs adjustment.

Continue with