OpenAI announced the launch of its new voice model GPT-Live-1 to power voice conversations in ChatGPT, aiming to make interactions with its intelligent assistant feel more like talking to a real person, with improvements including fewer interruptions, better understanding of natural pauses in speech, and smoother responses.

The company stated that the new model is its most advanced voice model to date; it can automatically utilize its best text models, such as GPT-5.5, when reasoning or web search is needed, then seamlessly return to the conversation to display results without the user feeling a break in the dialogue.

The GPT-Live-1 model also allows displaying AI-generated visual information during conversations related to weather, stocks, or sports, such as match results or weekly weather forecasts, providing a more interactive experience.

Previous versions of voice conversations relied on a turn-based model, which sometimes led to less accurate answers or less natural dialogue. The new model, however, is built on a structure called Full Duplex, enabling it to listen and speak simultaneously, processing inputs and generating responses concurrently and continuously.

This structure opens the door to new features, including real-time translation during speech; users no longer need to stop talking for ChatGPT to start translating—the conversation is translated moment by moment. It also allows users to ask the assistant to remain silent until called again, with cues indicating it is still listening, such as short acknowledgments.

OpenAI confirmed the model includes built-in safeguards to limit harmful responses and may end the conversation in cases deemed high-risk. It was also trained to provide expert-approved guidance when conversations about self-harm are detected, in addition to offering age-appropriate responses for younger users.

OpenAI has started rolling out GPT-Live-1 to ChatGPT users on iOS, Android, and the web. It will be available to subscribers of the Go, Plus, and Pro plans, while free account users will get a lighter version called GPT-Live-1 mini.

Introducing GPT-Live, a new generation of voice models for natural human-AI interaction.

Rolling out in ChatGPT starting today.

You’ll want to turn the sound on for this one. pic.twitter.com/WzoQFvA5ir

-- OpenAI (@OpenAI) July 8, 2026

Tags

ChatGPT OpenAI Artificial Intelligence

Link copied

July 9, 2026 | Last updated: July 9, 2026

1 minute

Follow us