OpenAI Launches GPT-Live With More Natural Conversations and Smarter ChatGPT Voice [Video]

OpenAI has unveiled GPT-Live, a new generation of voice models designed to make conversations with ChatGPT feel much more like a real conversation. Rolling out globally today, the system introduces a full-duplex architecture that can listen and speak at the same time, rather than waiting for each person to finish before responding, making conversations feel more natural.



Unlike previous turn-based or cascaded models, GPT-Live continuously processes input while generating output, deciding many times per second whether to speak, stay silent, pause, interrupt, or continue listening. OpenAI says this allows the assistant to naturally acknowledge what you're saying with phrases like "mhmm" or "got it," wait when you pause instead of jumping in too quickly, and better focus on your voice in noisy environments. The company has also remastered all nine built-in ChatGPT voices for the new system.


GPT-Live is also OpenAI's smartest voice model to date. For complex questions requiring web search or deeper reasoning, it delegates tasks to the latest frontier model, currently GPT-5.5, behind the scenes and brings the results back into the conversation when ready. While a task runs, the model can continue a free-flowing conversation. At launch, GPT-Live-1 will become the default voice model for Go, Plus, and Pro subscribers, while the smaller GPT-Live-1 mini will power ChatGPT Voice for free users.

The new experience also supports rich visual cards for weather, stocks, and sports, can perform live translation, and continues to support search, memory, images, and file uploads during conversations. OpenAI says GPT-Live is designed for conversation, not voice impersonation, and includes dedicated safeguards for teen users, protections against harmful content, and systems that can intervene in real time if unsafe responses are detected.

The rollout arrives as Apple has been expanding support for third-party voice assistants. Apple added CarPlay support for conversational AI apps like ChatGPT and Gemini in iOS 26.4, allowing those apps to integrate with the in-car interface. At launch, GPT-Live does not yet support voice with video or screen sharing in ChatGPT, though those features remain available in the legacy Standard and Advanced Voice modes while OpenAI works to bring them to GPT-Live.

Check out the videos below for more details...