OpenAI has launched GPT-Live, a new generation of voice AI models designed to make conversations with ChatGPT feel more natural. Announcing the new model, the company said that GPT-Live can listen and speak at the same time instead of waiting for a user to finish talking before responding. It also allows ChatGPT to continue a conversation while carrying out tasks such as web searches or deeper reasoning in the background. OpenAI said that the new GPT-Live voice model is rolling out globally from today on iOS, Android and ChatGPT.com. GPT-Live-1 will be the default voice model for Go, Plus and Pro users, while Free users will get GPT-Live-1 mini.GPT-Live brings faster and more natural conversationsOpenAI said GPT-Live is built using a full-duplex architecture. This allows the AI model to listen and speak at the same time. During a conversation, the model can respond with phrases such as "mhmm" or "yeah," pause when users are thinking, or interrupt naturally when needed. It can also keep talking while another AI model handles web searches or more complex reasoning in the background. At launch, GPT-Live uses GPT-5.5 for those advanced tasks.The company said GPT-Live is designed to replace the older turn-based voice system, where the AI had to wait until a user stopped speaking before replying. This reduces delays and makes conversations feel more like talking to another person.New features for ChatGPT Voice usersOpenAI said ChatGPT Voice now supports smarter answers, better listening and visual cards for topics such as weather, stocks and sports.Users can interrupt the AI, ask it to slow down or request that it simply listen without responding immediately. The company also said the model is better at ignoring background noise and focusing on the speaker's voice.GPT-Live also includes new safety features. OpenAI said the system can detect potentially unsafe conversations, guide users toward safer responses and provide crisis support information in cases involving self-harm. The company added that additional protections have been built for teenage users and that safeguards prevent the model from imitating the voice of a real person.At launch, GPT-Live does not support voice conversations with video or screen sharing, though OpenAI said those features are planned for a future update.