Google has officially released the Gemini 3 Pro AI model. Although it is currently in the preview phase and may take a few days to reach all regions, the company has already begun integrating it into its products. The first platforms to receive it are the Gemini app and the AI Mode in Google Search. The biggest new addition is generative interfaces, which make information discovery more visual and interactive.
What’s New in the Gemini App?
According to a blog post by Josh Woodward, VP of Google Labs, Gemini, and AI Studio, several updates are coming to the Gemini app on Android, iOS, and the web. Key improvements include:
- A new, modern interface that makes starting chats easier
- A My Stuff folder where users can find their created images, videos, and reports
- Shopping Graph integration for a smoother shopping experience
The most significant upgrade is generative interfaces (generative UI)—an interface that can automatically create personalized visuals, interactive elements, pages, and tools based on any prompt. Two experimental features are being introduced:
- Visual Layout
A magazine-style layout featuring photos and modules. For example, if you request a trip plan, you may see:
- Tap-able trip carousels
- A slider to choose trip duration
- Bars to select points of interest
Based on your selections, the AI creates a personalized plan without requiring detailed typing.
- Dynamic View
Using Gemini 3’s agentic coding abilities, the AI can design and code a custom interface in real time according to your prompt. For instance, asking, “Explain the Van Gogh Gallery with life context for each artwork” will generate an interactive window where you can click, scroll, and visually explore the content.
These features are rolling out now, but during initial testing, users may see only one of them.
Google is also introducing a new experimental feature called Gemini Agent, which can perform multi-step tasks and connect with Google apps to:
- Manage your Calendar
- Set reminders
- Draft emails and organize the inbox
- Handle web-based tasks like bookings and appointments
This feature is currently available on the web for AI Ultra subscribers in the US.
What’s New in AI Mode (Search)?
Google Search’s AI Mode is also being upgraded with Gemini 3 Pro. In the first phase, it will be available to AI Pro and AI Ultra subscribers in the US. Users can select “Thinking” from the model dropdown to try it. There will be initial usage limits, which will gradually increase.
Key enhancements include:
- Improved reasoning for complex queries
- Better ability to extract accurate responses from large numbers of web pages
- A new automatic model selector that routes difficult questions to Gemini 3 Pro and uses faster models for simpler ones
Generative interfaces are also arriving in AI Mode. This means the AI can instantly create custom visual layouts, tools, or simulations based on your prompt.
For example, when comparing mortgage loans, Gemini 3 AI Mode can create an interactive loan calculator, making it easier to compare options visually.