Introducing GPT-Live
OpenAI, Wednesday, July 8th, 2026
OpenAI releases GPT-Live-1 and GPT-Live-1 mini, full-duplex voice models for more natural real-time conversation.
OpenAI announced GPT-Live, releasing GPT-Live-1 and GPT-Live-1 mini for real-time voice interaction. The models support full-duplex operation, letting them listen and speak simultaneously for more natural, human-like conversations.
This removes the rigid turn-taking of prior voice systems, enabling smoother interruptions and back-and-forth dialogue. The mini variant offers a lower-cost option for latency-sensitive and high-volume voice applications.