What is Intelligent UI?
OpenAI introduced Intelligent UI in ChatGPT on October 7, 2026. It lets responses combine text, visuals, and interactive elements such as charts, buttons, and forms according to the question.
Intelligent UI vs. Generative UI
Generative UI describes a broader approach in which AI generates or composes interfaces for a user’s context. On this page, Intelligent UI refers to the specific ChatGPT capability named by OpenAI.
We include ChatGPT Intelligent UI as a Generative UI product case. Related products and frameworks below illustrate different approaches, each with its own capabilities.
Intelligent UI examples
These examples come from OpenAI’s announcement and show how a response can become something you interact with.
Interactive explanations
Explore a diagram or change an input to see how a concept works, rather than only reading an explanation.
Tools for a specific task
Ask for a bill splitter, savings calculator, or a game you can use directly in the conversation.
Visual planning
View a road trip on a map or a cooking schedule alongside a recipe, with details organized around your task.
ChatGPT and related Generative UI cases
Start with ChatGPT Intelligent UI, then explore other products that generate or present interactive interfaces.
ChatGPT Intelligent UI: Interactive Answers
OpenAI introduced Intelligent UI in ChatGPT on October 7, 2026. GPT-6 composes answers from text, visuals and interactive components, selecting a layout for the question. Official demonstrations include learning diagrams, trip planning, a bill splitter, a savings calculator and a game.
Copilot in SharePoint: Live Generated Dashboards
Microsoft's August 2026 SharePoint update describes Copilot generating interactive HTML dashboards from natural-language requests. The resulting views stay connected to SharePoint lists and refresh when opened; Excel and CSV can also supply dashboard data. Examples include project health, inventory, recruiting and approval pipelines.
Google Gemini 3 - Dynamic View & Visual Layout
Google has implemented Generative UI features in Gemini 3 (internally called "Dynamic View" and "Visual Layout" experiments). According to Google's official blog, Gemini 3 AI Mode can "generate immersive visual layouts and interactive tools and simulations", completely generating interfaces on-the-fly based on user queries. Dynamic View allows Gemini 3 to "design and code a fully custom interactive response" for each prompt, while Visual Layout generates "magazine-style" immersive multimedia interfaces.
Claude Artifacts: Shareable AI-Powered Apps
Anthropic's July 2025 Claude Artifacts update turns the artifact panel into a lightweight app surface. Claude can create artifacts that call Claude through an API, host them, and make them shareable while end users authenticate with their own Claude accounts. It is a useful case because generated UI becomes persistent, editable, and usable by other people, not just a one-off preview beside a chat.
Resources for building Generative UI
Explore structured UI output, component rendering, and streaming interfaces in these independent developer resources. They offer practical approaches for your own applications.
OpenAI Structured Outputs Sample: Generative UI
Official OpenAI sample app showing a practical structured-output path to Generative UI. The demo uses a recursive component schema plus a `generate_ui` tool so model responses can stream reliable UI definitions instead of only text, while still supporting multi-turn conversations and automatic tool execution.
vercel-labs/json-render
Vercel Labs' official open-source Generative UI framework. json-render lets models generate interfaces from prompts while staying constrained to a developer-defined component catalog, then progressively renders the resulting spec across web, mobile, and other runtimes.
assistant-ui: Component Vocabulary for Generative UI
Official implementation guide for @assistant-ui/react-generative-ui, an open-source runtime that lets models compose JSON interface trees through a present tool. It covers registered React components, Zod prop schemas, opt-in streaming, action handlers, and adapters for Slack, Teams, and A2UI.
Common questions
How can I try Intelligent UI?
In a supported ChatGPT experience, describe the task you want to explore, such as a visual explanation or a calculator. The response format depends on the question. Check OpenAI’s announcement for rollout information.
Are these frameworks the ChatGPT Intelligent UI SDK?
The announcement describes a ChatGPT capability. The frameworks listed here are separate Generative UI tools; their inclusion does not imply access to ChatGPT’s internal component library.
How does this differ from a traditional interface?
A traditional interface usually offers predefined screens and controls. A generated interface can compose a view around the current request, while still relying on components and interaction rules defined by its builders.