Introducing next-generation audio models in the API
OpenAI Blog
OpenAI released updated audio models in its API that allow developers to specify speaking styles for text-to-speech output. The model can now accept instructions such as "talk like a sympathetic customer service agent" to customize vocal delivery. This enables developers to create voice agents with more personalized communication patterns than previously available.
Why it matters
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.