GPT-Live-1

GPT-Live-1 is a full-duplex voice model designed for the OpenAI API that enables natural, real-time conversational experiences. It allows developers to build voice agents capable of simultaneous listening and speaking, featuring advanced interruption handling and customizable tone, pace, and style.

GPT-Live-1 is a full-duplex voice model designed for the OpenAI API that enables natural, real-time conversational experiences. It allows developers to build voice agents capable of simultaneous listening and speaking, featuring advanced interruption handling and customizable tone, pace, and style.

• 01Full-duplex voice interaction allowing simultaneous listening and speaking
• 02Advanced interruption handling that reduces latency and improves conversational flow
• 03Flexible delegation of reasoning and tool calling to backend models like GPT-6 Astra
• 04Customizable agent tone, pace, and conversational style via system prompts
• 05Native telephony support for real-time voice agent deployment
• 06Improved context retention and background noise management for long-session reliability