OpenAI Launches GPT-Live: ChatGPT Can Now Listen and Speak at Once

OpenAI has officially launched its new GPT-Live voice model, designed to make interacting with AI feel like a genuine human conversation. The company began rolling out the update today across the ChatGPT mobile apps for iOS and Android, as well as the web interface.

A key feature of GPT-Live is its full-duplex architecture. This allows the AI to listen and speak at the same time, enabling more natural back-and-forth exchanges, better timing, and the ability to perform live translations. OpenAI says it has also given the model a more defined and enjoyable personality compared to previous versions.

Behind the scenes, the system is designed to delegate complex tasks. This includes things like web searches or deep reasoning, using OpenAI’s most advanced models. Once the model gathers the necessary information, it then feeds the answer back into the conversation in real time.

OpenAI CEO Sam Altman said today, “GPT-live (next-generation voice) launches today in ChatGPT. it feels magical and ‘real’. i have always preferred typing to talking to an AI, now i think that’s going to shift.”

OpenAI expects the rollout to reach all users over the next few days. To access the new feature, users can simply tap the voice button within the app. GPT-Live will also be coming to the OpenAI API in the near future.