Intelligent UI

Intelligent UI is the name OpenAI gave to a capability shipped with GPT-6 in ChatGPT on October 7, 2026. Instead of the model returning a wall of text every time, the system can compose a response from native, streamable UI components: a bar chart if you asked a data question, a form if you need to fill something in, tappable buttons if there are clear next steps, or a small interactive calculator if numbers are involved. The model decides which components fit the question, not the developer.

The underlying mechanism is a component library paired with a compiler that renders elements progressively as the model generates its answer. That means you can start reading or interacting before the full response is finished, similar to how streaming text works today but extended to interactive elements. OpenAI described the goal as software adapting to people rather than people adapting to software.

For builders, Intelligent UI is relevant as a signal about where the AI interface layer is heading. The idea that a model can own layout and interaction decisions, not just content, connects to earlier concepts like generative UI and dynamic artifacts. It also raises new design questions: if the model picks the format, how do you build consistent product experiences on top of it?