↗← All demos POST /v1/audio/speech

08 · Text to speech

Text to speech

Turn text into spoken audio. With gpt-4o-mini-tts you can also describe how the voice should sound: tone, speed, emotion or accent. The API sends back an audio file, not JSON. Docs ↗

Voice
Marin and Cedar are the newest, most natural voices. Voices that are crossed out do not work with the selected model.
Only gpt-4o-mini-tts follows these instructions.