Tag
The latest update to OpenAI’s audio stack signals a shift from simple text-to-speech toward low-latency, agentic conversation.