Gemini 3 Release Comes With a Redesigned Gemini App, New Features for AI Mode

Google has now released the Gemini 3 Pro artificial intelligence (AI) model. Although it is available in preview, and the rollout across the globe might take a couple of days, the Mountain View-based tech giant has already started announcing its integration across its suite of products. Two of the company's products that are getting the model's capabilities are the Gemini app and AI Mode in Search. The tech is introducing a new capability called generative interfaces in both of these platforms, which will offer users a more visual and interactive way of finding information.

New Features in Gemini App

In a blog post, Josh Woodward, the Vice President of Google Labs, Gemini and AI Studio at Google, introduced new improvements in the Gemini app. While the company uses the phrase “app,” these changes will be available across the Android and iOS app, as well as the Gemini web interface.

The company says it is redesigning the Gemini platform with a clean and modern look. What that means is users will find it easer to start chats; find images, videos, and reports they have created via a new My Stuff folder; and enjoy shopping for products with the integration of Google's Shopping Graph.

However, the biggest introduction is generative interfaces or generative UI. In a separate blog post, the company describes this capability as an interface “which dynamically creates immersive visual experiences and interactive interfaces — such as web pages, games, tools, and applications — that are automatically designed and fully customized in response to any question, instruction, or prompt.”

The Gemini app is first releasing generative interfaces with two experiments. First is visual layout. It is a magazine-style view that features photos and modules. So, if a user prompts the chatbot for a itinerary for a trip, Gemini might show a carousel of different types of trips that users can tap on to select, a slider might let them select how many days they will be on a vacation, and another bar might let them click on different point-of-interests. The selection will then allow the AI to curate the right plan, without them having to spell it out by either typing or speaking.

Second is dynamic view. Using Gemini 3's agentic coding capabilities, the chatbot can design and code a custom user interface in real-time in response to a user prompt. For instance, if a user asks Gemini to “explain the Van Gogh Gallery with life context for each piece,” it will generate an interactive window where users can click on elements, scroll, and slide to learn about the topic visually.

Both of these features are rolling out now, although the company says users will initially only see one of them to help the company compare between the two.

Finally, the company is also introducing Gemini Agent. It is an experimental feature that can perform multi-step tasks within the app. However, it can connect to Google apps to manage the user's Calendar, add reminders, and even organise their inbox by drafting replies to emails. It can also perform web-based tasks such as making bookings or taking appointments. It is currently available on the web for Google AI Ultra subscribers in the US.

New Features in AI Mode

AI Mode in Search is also getting a few new capabilities with the integration of Gemini 3 Pro. However, these are first rolling out to the Google AI Pro and AI Ultra subscribers in the US. They can now select “Thinking” from the model dropdown menu to access the latest AI model. Currently, the rate limit with Gemini 3 Pro will be limited, but the company states that it will be increased soon.

So, what's new in AI Mode? Due to improved reasoning capability, AI Mode can now tackle more complex queries and intelligently sift through a large number of web pages to find the contextually relevant responses. A new automatic model selection tool is also being added to Search, which will route user's more challenging questions in AI Mode and AI Overviews directly to Gemini 3 Pro. For simpler questions, it will continue to use the faster models.

Generative interfaces are also making their debut in AI Mode. Google says this will allow the AI tool to create visual layouts for responses in real-time, complete with interactive tools and simulations based on user queries.

“When the model detects that an interactive tool will help you better understand the topic, it uses its generative capabilities to code a custom simulation or tool in real-time and adds it into your response,” the post explained.

Highlighting an example, the tech giant said if a user is researching mortgage loans, Gemini 3 in AI Mode can create a custom interactive loan calculator directly in the interface to help them compare different options.