OpenAI launched GPT‑Live‑1 in its API, bringing full-duplex voice conversations—simultaneous listening and speaking—to developers. Introduced earlier in ChatGPT, the model can delegate reasoning and tool calls to backend models, as seen with Codex and ChatGPT Work. In early evaluations, Speak reported nearly 80% fewer interruptions versus previous turn-based systems. OpenAI says GPT‑Live‑1 improves Full Duplex Bench performance by 30 percentage points over GPT‑Realtime‑2.1 and supports telephony deployments.
No score is assigned. Sources and their independence are shown in the citation chain below.