OpenAI has launched GPT-Live, an innovative voice AI system designed to enhance conversational experiences by enabling simultaneous speaking and listening. This advanced AI is built on a full-duplex architecture, allowing it to engage in real-time dialogue, offering natural acknowledgments like “mhmm” or “yeah” to facilitate smoother interaction without awkward pauses.
When handling more complex commands that require tasks such as web searches or intricate reasoning, GPT-Live can effortlessly delegate these to a more advanced AI model operating in the background, while maintaining the conversation flow. Currently, these sophisticated tasks are managed by GPT-5.5, with plans to integrate more advanced frontier models in upcoming updates.
The introduction of GPT-Live follows OpenAI’s recent announcement regarding its latest model series, GPT-5.6, which is set to be released publicly after further cybersecurity assessments. The GPT-5.6 family features the flagship Sol model, along with Terra and Luna variants, expanding OpenAI’s reach in the field of AI technology.
OpenAI has begun deploying two versions of the new voice models—GPT-Live-1 and GPT-Live-1 mini—to ChatGPT users across the globe. Additionally, the company aims to extend the availability of GPT-Live through its API, allowing developers and businesses to incorporate real-time voice AI capabilities into their applications.