ChatGPT’s Gemini Live opponent can keep talking while thinking

in interaction with AI Talking to models has been possible for a while, but it’s never been the most seamless experience. However, OpenAI’s latest models may change that for the better.

OpenAI exists today announced its new GPT-Live sound model. It is built on a full-duplex architecture, which means the model can listen and talk to users at the same time. This can result in more natural conversations with GPT-Live. According to the report, the model will be able to show attention by using phrases like “mhmm” or “yes” while the user is speaking.

To do all this, GPT-Live continuously processes inputs and decides to talk, listen or pause while listening to the user. This way, the user can interrupt with natural responses even while they are still talking. A chat sample shared by OpenAI showed that GPT-Live interacts with the user more naturally than before.

GPT-Live also takes a new approach to questions that require more thinking or web searching. The model basically assigns the task to a model like GPT-5.5 and continues to communicate with the user. ChatGPT is a neat way to ensure the chat doesn’t get stuck while doing background research.

GPT-Live users are also generally better at listening. If the user pauses to think while speaking, GPT-Live will wait for the user to finish speaking. Users will also be able to ask ChatGPT to be quiet and listen, and better focus on the user’s voice without being distracted by the surrounding noise.

The new model also supports visual responses. It can display rich visual cards for questions about weather, stocks, sports, and more.

If you have used Gemini Live in the past, OpenAI’s new model might have looked quite similar. However, GPT-Live seems more advanced than Google’s offering. Gemini Live can keep up with long conversations, but it’s not as responsive as the GPT-Live demos. GPT-Live can also render visual results and delegate tasks to other OpenAI models to keep the conversation smooth and natural, something Gemini Live can’t yet do. This may change in the future, but so far GPT-Live seems to have the upper hand.

OpenAI releases two models of GPT-Live: GPT-Live-1 and GPT-Live-mini. Both models go on sale today. GPT-Live-1 will be the standard practice ChatGPT ProPlus and Go users, and GPT-Live-mini will be the default model for free users. The new voice experience will be available on ChatGPT’s smartphone apps as well as on the web.