Google has unveiled extensive AI features covering its search system, upgrading the Gemini app to function more as a work assistant.
Google revealed a major set of artificial intelligence technologies at Google I/O 2026, led by the Gemini 3.5 Flash model, which delivers output tokens per second four times faster than other models. They also introduced Gemini Omni, a new series model that accepts inputs including images, audio, video, and text, then produces video output. Additionally, the Gemini application was upgraded with a new design language called Neural Expressive, featuring smooth animations and an easier-to-use interface.
A key highlight of Google I/O 2026 was the launch of Gemini Spark, a personal assistant agent that can autonomously perform tasks on behalf of users by connecting with Gmail, Docs, and other Google Workspace apps. Alongside this, the Daily Brief feature summarizes and prioritizes daily tasks by pulling data from Gmail, Calendar, and Tasks, while also suggesting next steps for the user.
Google Search, the well-known search system, received an AI mode upgrade featuring Gemini 3.5 Flash, which includes an intelligent search channel and a 24/7 background search agent that continuously updates news and real-time information tailored to individual interests.
There is also the Universal Cart feature, a smart shopping cart that helps find deals, notifies users of discounts, displays historical price data, alerts stock status, and checks hardware component compatibility when shopping across various Google platforms.
For professional users, new features include Gmail Live, enabling real-time conversational search within emails, and Docs Live, which allows document creation and editing through voice commands.
On the hardware side, Google announced smart glasses under the Android XR platform, set for release late this year. This involves partnerships with Samsung and Qualcomm for hardware development, and collaborations with fashion brands Gentle Monster and Warby Parker on design. The glasses will support connectivity with both Android and iPhone devices. Alongside this, the Android Halo feature was introduced to display the AI agent’s activity status at the top edge of smartphone screens, letting users know what the AI is doing without switching screens.
Furthermore, Google has restructured its Google AI service billing, moving from daily usage quotas to a model based on processing volume to better reflect the varying complexity of different command types.