OpenAI launches GPT-Live-1 for ChatGPT Voice

OpenAI has launched GPT-Live-1, introducing a new voice experience in ChatGPT designed to make conversations feel more natural and responsive. The company is rolling out GPT-Live-1 for paid users and GPT-Live-1 mini for Free users.

The new models can listen and speak simultaneously, allowing users to interrupt, pause or continue speaking while ChatGPT responds. OpenAI says this improves turn-taking and makes voice interactions feel closer to a natural conversation.

GPT-Live-1 works within a standard ChatGPT conversation, with spoken responses appearing alongside streamed text. The model can also use web search and memory, display visual results through supported widgets, and work with text and images where those features are available.

OpenAI says GPT-Live is rolling out globally on ChatGPT.com and the ChatGPT iOS and Android apps. GPT-Live-1 will become the default voice model for Go, Plus and Pro users, while GPT-Live-1 mini will serve as the default for Free users.

At launch, GPT-Live is not available in ChatGPT Business, Enterprise or Edu workspaces. It also does not currently support video or screen sharing, although eligible users can continue using those features through Advanced Voice Mode where available.

OpenAI says GPT-Live-1 can hand more complex tasks to other models, such as GPT-5.5, when they require search, advanced reasoning or more agentic capabilities. The company also plans to make GPT-Live available through its API in the future.

Why does it matter?

GPT-Live-1 reflects OpenAI’s broader effort to make voice a core interface for interacting with AI rather than a separate feature. By combining real-time speech, streamed text, search, memory and visual results within a single conversation, the company is moving towards more seamless multimodal assistants capable of supporting everyday tasks, research and longer, more natural interactions.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!