Beating AI News Flash: OpenAI has opened the GPT-Live-1 API, the real-time voice model currently used by ChatGPT. It can listen and speak simultaneously, directly understand speech, handle ordinary conversations, and also cope with interruptions, pauses, and background noise.
When encountering complex reasoning or tool calls, GPT-Live-1 can hand off tasks to GPT-6 Astra, Luna, or third-party models, then continue chatting with the user after receiving the results. Developers can decide for themselves what model to connect on the backend.
Compared with the traditional "speech-to-text → large model → text-to-speech" process, GPT-Live-1 reduces intermediate model handoffs, making real-time conversations more coherent.
GPT-Live-1 is priced at $0.05 per minute, with backend models billed separately.

