ChatGPT’s new voice model promises more natural and human conversations By Maria
2 min read
 |
OpenAI unveiled on Thursday, July 9, 2026, a new artificial intelligence model, GPT-Live, designed to make ChatGPT voice interaction more natural and seamless. This update allows the system to listen while the user speaks, responding in a more fluid and direct way. The main innovation is the ability to process audio and generate speech simultaneously, which transforms the dynamics of conversations. The GPT-Live-1 and GPT-Live-mini versions are now accessible from this date.
How the new model transforms voice interaction
GPT-Live uses a “full-duplex” architecture, an advancement that allows the system to process the user’s audio continuously while they express themselves. This capability allows responses to be issued in a more integrated and brief manner, only when appropriate. Furthermore, the technology allows AI to detect pauses in speech, waiting for the user’s idea to be completed before intervening, which is much closer to a real human conversation.
Introducing GPT-Live, a new generation of voice models for natural human-AI interaction.
— OpenAI (@OpenAI) July 8, 2026
Rolling out in ChatGPT starting today.
You’ll want to turn the sound on for this one. pic.twitter.com/WzoQFvA5ir
One of the notable features is the model’s ability to sustain the flow of the conversation, even when faced with complex questions that require greater reasoning or web searching. During these times, GPT-Live continues to interact with the user while the request is forwarded to another advanced model for processing. This mechanism was implemented to significantly reduce downtime and waiting during voice interactions.
Images and visual data will complement conversations
GPT-Live will not be restricted to just auditory communication, promising to go further. Soon, the platform will be able to present “cards” with visual information on a variety of topics, such as weather forecasts, financial market data and the results of sporting events. This functionality aims to enrich the ChatGPT voice experience, offering an interactive visual complement directly on the screen.
Details about the availability of new models
OpenAI confirmed that the GPT-Live-1 model will be the default voice experience for subscribers of the ChatGPT Pro, Plus and Go versions. In turn, free users will have access to the GPT-Live-mini model. The new voice functionality will be implemented and accessible on both mobile apps and the web platform.
