OpenAI released a demo video on July 15 showing GPT-Live handling several tasks at once while keeping a continuous voice conversation with the user. In the demonstration, the model checks flights, pulls local weather and drafts an itinerary in real time.

The video offers a concrete look at how GPT-Live is meant to be used. The product went live on July 8 as OpenAI’s full-duplex voice model. In a post on X, OpenAI said the core of this multitasking capability is that the model “doesn’t have to wait for the user to finish speaking before responding.”
Full-duplex design lets the model listen and speak at the same time
According to OpenAI’s official announcement, GPT-Live formally replaced ChatGPT’s previous Advanced Voice Mode on July 8. The defining feature of the system is its full-duplex architecture, which allows it to take in user speech while also generating spoken output, making real-time conversation possible.
During a conversation, GPT-Live can insert short acknowledgments such as “mhmm” and “yeah” to signal that it is still following along. It can also stay silent when the user needs time to think, instead of forcing a reply into the exchange.
Two versions are available, with GPT-5.5 handling deeper reasoning and search
GPT-Live is available in two versions. GPT-Live-1 is the default for paid plans, while GPT-Live-1-mini is offered on the free tier.
When a conversation requires deeper reasoning or search, GPT-Live can delegate that part of the task to the more capable GPT-5.5 while continuing to keep the live conversation flowing. OpenAI’s July 15 demo of checking flights, weather and itinerary details at the same time is built on that mechanism.
Screen sharing and camera support are still missing
OpenAI said GPT-Live still does not support screen sharing or camera use. Those features remain on the older ChatGPT Voice Mode. The company has not given a timeline for adding visual input to GPT-Live.

OpenAI says both GPT-Live versions beat the older Advanced Voice Mode in tests
In comparison tests lasting five to 10 minutes, both GPT-Live-1 and GPT-Live-1-mini performed better than the previous Advanced Voice Mode on conversation preference, turn-taking, interruption handling, fluency and the natural feel of the interaction.
GPT-Live is now available globally on iOS, Android and the web, where it has replaced the older Advanced Voice Mode.

