Generative UI Is One Of Gemini 3 Pro's Best New Features

By Chris Smith
The Gemini 3 logo. Google/YouTube

Google unveiled the Gemini 3 models on Tuesday, shortly after the Gemini 3 Pro benchmarks and model card leaked, teasing massive performance gains over rival products from OpenAI and Anthropic. The benchmark scores could not detail all the novelties in Gemini 3 Pro, especially the new Generative UI functionality, a feature that's not available from competing AI services like ChatGPT. Generative UI lets Gemini create rich visuals when crafting responses to user prompts. The visuals can help explain a topic better than text-only responses. The new feature can be used to turn a complex concept into something simpler or to create a response that looks like a website or app experience.

The new Gemini 3 model powers Generative UI, creating the advanced visuals and UI on the spot, as it generates the response and determines whether UI visuals are needed. The resulting UI can include interactive elements, like buttons and menus. Also, Gemini can create immersive web, app, and game experiences within its reply.

Google explained in a Google Research blog post that Generative UI may be the precursor of something much bigger. "This work represents a first step toward fully AI-generated user experiences, where users automatically get dynamic interfaces tailored to their needs, rather than having to select from an existing catalog of applications," Google said, teasing a world where people may use AI to create bespoke apps on the fly.

Generative UI uses several "key tools," including an unidentified image generation model and access to web search. Generative UI will work initially in the Gemini app (the Dynamic View and Visual Layout experiments) and in Google Search (AI Mode Thinking model), though the latter will be available only to Google AI Pro and Ultra subscribers in the U.S.

What can Gemini 3 do with Generative UI?

Google offered a few examples of Generative UI responses. For example, the following video shows Gemini 3 responding to this prompt: "Create a Van Gogh gallery with life context for each piece." The AI creates a website experience featuring an interactive menu that can be clicked to access different paintings and relevant information.

The following example shows how Gemini 3's Generative UI might help students struggling to understand complex subjects, or regular Google Search users who might be curious about a specific topic. Answering this prompt, "show me how RNA polymerase works. What are the stages of transcription and how is it different in prokaryotic and eukaryotic cells," AI Mode creates an animated visual that illustrates the chemical reactions via a simulator accompanying text generated by the AI.

Google offers additional examples of what's possible with Generative UI in Gemini 3 in the Google Research blog post. The AI created a website to showcase different outfit styles for a user who sent the following prompt: "Be my fashion advisor — explain the theory and show me a dozen stylish options for me." The website contains realistic images, menus, and descriptions similar to what you might see in an online store dedicated to fashion. The difference is that the website has been created within Gemini experiences for a single person.

An example of a Generative UI response in Gemini 3.Google

Gemini will also be able to tailor the Generative UI design to the content consumer. A prompt asking for a math experience for a 5-year-old ("Teach my 5-year-old addition, subtraction, multiplication, and binary representation with cute creatures that play basketball. He loves basketball") leads to a Generative UI response that features an interactive basketball-themed website the boy can use to learn math.

An example of a Generative UI response in Gemini 3.

Recommended