OpenAI has launched GPT-Live voice models, which enable simultaneous listening and speaking during conversations, enhancing interaction quality. This new technology powers an updated ChatGPT Voice experience available to consumers in supported regions and integrates features like web search and memory. The full-duplex audio processing capabilities of these models allow for real-time responses, interruptions, and natural turn-taking, significantly advancing the functionality of voice AI.

OpenAI: OpenAI develops generative artificial intelligence models and applications, including the GPT series and the ChatGPT platform. The company has introduced GPT-Live voice models designed for simultaneous listening and speaking to create more fluid interactions in its voice-enabled products.

Consumer Access: The GPT-Live-1 model powers an updated ChatGPT Voice experience rolling out across consumer plans in supported regions, with integration for web search and memory features.
Product Integration: The new voice capabilities support simultaneous audio and text handling along with visual result displays in conversations.
Voice AI Advancement: OpenAI’s GPT-Live models enable full-duplex audio processing, allowing real-time responses, interruptions, and natural turn-taking during conversations.