Google Unveils Major Gemini App Enhancements at I/O 2026
At the recent I/O 2026 conference, Google introduced significant updates to its Gemini app, reflecting the company’s commitment to advancing artificial intelligence and enhancing user experience. With over 900 million monthly active users, the Gemini app has become a central hub for AI-driven interactions.
Neural Expressive Redesign
The Gemini app has undergone a comprehensive redesign, adopting the Neural Expressive design language. This new aesthetic emphasizes fluid animations, vibrant colors, haptic feedback, and updated typography to create a more engaging user interface.
Key changes include a pill-shaped prompt box that simplifies user input. On mobile devices, a singular ‘plus’ menu opens a bottom sheet featuring a carousel of upload options such as Photos, Camera, and recent images. Additional upload options like Files, Drive, and Notebooks are accessible by scrolling to More uploads. The Tools section now offers features like Guided Learning, Deep Research, Canvas, Create Music, Create Video, Create Image, and Personal Intelligence.
Voice input and Gemini Live have been streamlined for a more seamless experience. Gemini Live now operates inline, eliminating the need to switch between modes. The interface includes a centered waveform pill, with video and screen sharing options on the left, and mute and close buttons on the right.
The background now features a gradient with the Gemini logo and a greeting at the center. Temporary chats are accessible at the top-right corner, and the account switcher has been repositioned. The model picker is now a dropdown in the top-left corner.
The navigation drawer has been expanded to a fullscreen interface, displaying options like New Chat, Search Chats, Library (formerly My Stuff), and Gems. Notebooks and Recent are also included in this menu. At the bottom, users will find their profile picture and a settings gear for account management.
As part of the Neural Expressive initiative, Gemini’s responses are now more dynamic, presenting key information prominently and incorporating inline images, narrated videos, timelines, and interactive visualizations when relevant. This redesign is currently rolling out to Android, iOS, and web platforms.
Introduction of Gemini 3.5 Flash
Google has unveiled Gemini 3.5 Flash, a model that combines advanced intelligence with the capability to perform agentic tasks. This model surpasses Gemini 3.1 Pro in coding, agentic, and multimodal benchmarks, achieving:
– Terminal-Bench 2.1: 76.2%
– GDPval-AA: 1656 Elo
– MCP Atlas: 83.6%
– CharXiv Reasoning: 84.2%
Despite its enhanced capabilities, Gemini 3.5 Flash maintains the speed characteristic of the Flash series, operating four times faster than other frontier models in terms of output tokens per second. It also offers cost efficiency, being approximately one-third to one-half cheaper.
Gemini 3.5 Flash is now available in the Gemini app, Google Search, Google Antigravity 2.0, and the Gemini API. Additionally, Gemini 3.5 Pro is slated for release next month, focusing on similar advancements.
Launch of Gemini Omni
Google introduced Gemini Omni, a new model that integrates Gemini’s reasoning capabilities with creative functions. This model is designed to accept various inputs, including images, audio, video, and text, and can generate lifelike video content grounded in real-world knowledge. The generated videos are easily editable, catering to a wide range of creative applications.
Gemini Omni is currently rolling out to AI Plus, Pro, and Ultra users within the Gemini app, Google Flow, and YouTube Shorts. Furthermore, Google Flow and Flow Music are now available as mobile applications, with the Flow app in beta on Android (iOS version forthcoming) and Flow Music available on iOS (Android version forthcoming).
Introduction of Gemini Spark
Gemini Spark is a new feature within the Gemini app, acting as a personal agent to assist users in managing their digital lives. This feature represents a significant shift, transforming Gemini from a passive assistant into an active partner capable of performing tasks on behalf of the user.
Gemini Spark integrates seamlessly with Google Workspace applications such as Gmail and Docs, with plans to expand support to third-party tools via the Multi-Channel Platform (MCP) over the summer.
Additional Enhancements
The Gemini app now offers the ability to generate various file types directly within the application. Users can create and export files in formats including Google Docs, PDF, Word (DOCX), Excel (XLSX), CSV, LaTeX, Plain Text (TXT), Rich Text Format (RTF), and Markdown (MD). This functionality streamlines the process of converting ideas into shareable documents without leaving the app.
For example, users can prompt Gemini to Create an in-depth study guide as a PDF in LaTeX format using my lecture notes, including visuals, graphs, and equations. This feature is now available to all Gemini app users globally.
In terms of personalization, the Gemini app’s features like Memories, which allows Gemini to remember key details and preferences shared by the user, are expanding to the UK. Additionally, the app now supports importing chats from other AI applications, with full rollout expected in the coming weeks.
These updates underscore Google’s dedication to enhancing the Gemini app’s functionality and user experience, solidifying its position as a leading AI-powered assistant in the digital landscape.
