Gemini gets a brand new experience! It's already in Czechia, and this is what it looks like

At the I/O 2026 conference, Google introduced not only new models and smart glasses but also a major redesign of its flagship AI application. The Gemini app, which Google says is used monthly by over 900 million people in 230 countries worldwide, is getting a new user interface, new features, and new agents that are gradually taking over routine tasks. The goal is to shift Gemini from an assistant that answers questions to an assistant that directly gets things done for the user.

Neural Expressive: new visuals and way of conversing

The main visual novelty is the Neural Expressive design language. Google is replacing the existing Gemini interface with it, focusing on more vibrant colors, smoother animations, new typography, and haptic feedback during interactions. The app is intended to appear livelier and less static than before – Google describes it as a “design language for the AI era,” roughly translated from marketing speak as “it will move more.”

A more important change, however, concerns how you speak with the assistant. Gemini Live, previously a separate mode for voice conversation, is now integrated directly into the main environment. This allows you to seamlessly switch from typing text to spoken conversation and back again without interrupting the context. Google has also redesigned the microphone – it should better tolerate a slower pace of speech and not interrupt the speaker mid-sentence. Furthermore, regional dialects are planned, which will allow users to choose a specific voice variant. Google did not specify whether this will also apply to Czech and Slovak.

The responses themselves have also changed. Instead of long blocks of text, Gemini will more often respond using visual elements, interactive timelines, narrated videos, or dynamic graphics. It will depend on the type of query – for example, a request for historical context might return an interactive timeline. The new visuals and redesigned interface are rolling out globally starting today on the web, Android, and iOS.

Daily Brief as your morning coffee with AI

A new agent in the app is Daily Brief. This is a personalized morning brief designed to serve as the first interaction with your phone during the day. After enabling this feature, Gemini will scan connected applications (Gmail, Calendar, and others) in the background, select important events, unanswered emails, and upcoming deadlines, and compile a clear summary from them.

Furthermore, the agent will actively sort information by priority and suggest next steps – for example, “reply to this email” or “prepare materials for the 2:00 PM meeting.” The feature originated from a Google Labs experiment referred to as CC. Users can rate the outputs with a thumbs up or down, thereby “teaching” the agent their preferences over time.

Daily Brief is starting to roll out today, but currently only for subscribers of Google AI Plus, Pro, and Ultra plans in the United States. Google did not specify a global availability date.

Spark, Omni, and Gemini 3.5 in the app

The app also serves as a gateway to three other new features introduced at I/O 2026. Under the hood, the Gemini 3.5 Flash model is now running, which Google has deployed as the default. Subscribers to AI Plus and higher plans also have access to the generative video model Gemini Omni, which can create short clips directly within the app based on text, image, and video inputs, and then edit them with voice commands.

The third new feature is the personal agent Gemini Spark, designed to operate 24 hours a day and perform long-term tasks. Spark integrates with Google Workspace tools (Gmail, Docs, Slides), and also works with applications like Canva, OpenTable, or Instacart, with more partners gradually being added. Before any “risky” action – typically before spending money or sending an email – Spark is designed to request user confirmation. Spark is being rolled out to selected testers this week, with a beta coming next week for Google AI Ultra subscribers in the USA.

Gemini is also coming to macOS

A slightly surprising new development is the expansion of Gemini to Apple desktop. The macOS application is available for download starting today, but it does not yet include all features. A key expansion will come during the summer – Gemini Spark will arrive in the macOS app and will be able to work with local files and automate tasks across the desktop.

A new voice feature will also be added. Gemini is designed to generate accurate text from freely spoken speech – including various “um,” “like,” and misspoken words – format it according to the surrounding text’s context, and insert it precisely where the user’s cursor is. This feature is particularly useful for writing messages, emails, or meeting notes.

Do you like the direction Gemini is moving in?

Source: Google Blog