OpenAI releases new voice models for more natural live conversations
Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.
OpenAI's new GPT-Live-1 models enable full-duplex voice interactions, allowing simultaneous speaking and listening with natural interruption handling, replacing the previous three-step pipeline. This reduces latency and enables seamless live translation, making voice a viable primary interface for complex agentic workflows like coding or research. Engineers must now optimize for continuous, interruptible voice interactions rather than discrete turn-based exchanges.
OpenAI has launched GPT-Live-1 and GPT-Live-1 mini, full-duplex voice models that can simultaneously listen and speak to handle interruptions naturally while routing complex search, reasoning, and agentic queries to backend models like GPT-5.5. For production, this enables you to build hands-free, long-form voice agents that can process continuous, multi-minute context and dynamically switch to visual outputs without the high-latency bottleneck of traditional cascaded speech-to-text-to-speech pipelines.