Introducing next-generation audio models in the API

【RSS 原文 · 未经 AI 筛选或翻译】For the first time, developers can also instruct the text-to-speech model to speak in a specific way—for example, “talk like a sympathetic customer service agent”—unlocking a new level of customization for voice agents.