OpenAI launches GPT-Live voice models with real-time simultaneous listening and speaking

OpenAI has unveiled GPT-Live, a new family of voice models designed to enable artificial intelligence systems to listen and speak simultaneously in real time, marking its latest step toward more natural and responsive voice interactions.The company announced that two variants—GPT-Live-1 and GPT-Live-1 mini—will begin rolling out globally on Wednesday. The launch is aimed at improving the performance of voice-enabled AI applications by reducing interruptions and enabling conversations that more closely resemble human dialogue.Unlike conventional voice assistants that often wait for users to finish speaking before responding, GPT-Live is built to process speech continuously, allowing AI systems to understand, respond and adapt while a conversation is still underway. This capability is expected to improve the responsiveness of AI-powered assistants, customer support tools and other voice-driven applications.OpenAI is also preparing to expand the availability of its GPT-5.6 model after receiving approval from the U.S. Department of Commerce for a broader rollout, according to a report by Axios.Also read: Google unveils Gemini-powered marketing tools to boost AI-led advertising for Indian businessesThe approval follows additional testing and discussions between OpenAI and U.S. government officials. The report said the Department of Commerce's Center for AI Standards and Innovation evaluated the model, while OpenAI's technical teams remained in Washington to respond to questions and provide technical clarifications during the review process.The broader rollout is expected to begin later this week, marking a significant step beyond the model's initial limited release. GPT-5.6 was first introduced last month through a phased deployment that restricted access to a small group of trusted partners after the Trump administration requested a controlled rollout while it assessed the model's capabilities.At the time, OpenAI said the limited release was a temporary measure rather than its preferred launch strategy. The company maintained that delaying wider access slowed adoption among developers, businesses, cybersecurity professionals, researchers and global partners that could benefit from its latest AI capabilities.