GPT-4o Audio Blog

GPT-4o Audio API
Jan 6, 2026
GPT-4o Audio

GPT-4o Audio API

GPT-4o Audio API: A unified /chat/completions endpoint extension that accepts Opus-encoded audio (and text) inputs and returns synthesized speech or transcripts with configurable parameters (model=gpt-4o-audio-preview-, speed, temperature) for batch and streaming voice interactions.