Sundar Pichai at Google I/O 2026
Google spent its biggest developer event of the year arguing that AI should act, not just respond. At Google I/O 2026held May 19-20 in Mountain View, the company unveiled a new model family, a 24-hour personal agent and a smart shopping cart, along with what it called the most profound change to Search in the product’s history. CEO Sundar Pichai described the moment stage to the beginning of the Gemini agent era, and the word agent is attached to almost every product that appears.
The change matters now because of how quickly Google usage is growing. Pichai said the Gemini app has surpassed more than 900 million monthly users, up from 400 million a year earlier, with daily requests growing about sevenfold and API models now processing about 19 billion tokens per minute. He said Google expects capital spending of about $180 billion to $190 billion this year, up from $31 billion in 2022, much of which will go toward its eighth-generation TPU 8t and TPU 8i chips for training and inference.
New models lead the series
The front page was Gemini 3.5 Flashthe first model of the Gemini 3.5 family. It became generally available on launch day through the Gemini app, AI Mode in Search, Gemini API, Gemini Enterprise Agent Platform and Antigravity development platform. Google says the model outperforms the previous Gemini 3.1 Pro in demanding coding and agent tests, scoring 76.2% in Terminal-Bench 2.1 and 83.6% in MCP Atlas, while running about four times faster than other frontier models. A larger Gemini 3.5 Pro is already in internal use and is expected to be released next month.
Google also presented Gemini Omnia separate model designed to produce output from any input, starting with video. Google says the model combines the thinking of Gemini with a sharper understanding of physics, including gravity, kinetic energy and fluid dynamics, so the scenes created look more realistic. Videos created with Omni carry Google’s subtle SynthID watermark. Gemini Omni Flash is now available to all Google AI Plus, Pro and Ultra subscribers worldwide via the Gemini app and Google Flow, and is also free on YouTube Shorts Remix for users 18 and over.
Search is getting its biggest change in 25 years
Google called the new search box the biggest upgrade to this surface in 25 years. It allows users to search using text, images, files, videos and Chrome tabs, with search reasoning on all at once. The company is also folding AI insights and AI functionality into a seamless experience. AI Overviews now reaches more than 2.5 billion monthly users, and the AI feature, supported globally by Gemini 3.5 Flash as the default model, has surpassed 1 billion.
The biggest story is his arrival Search agents. Google starts with intelligence agents, which work in the background 24/7 to monitor a topic on the web and send synthetic updates. They’re rolling out this summer, first to Google AI Pro and Ultra subscribers. Search can also create custom layouts on the fly, a feature Google calls Generative UI, which arrives for everyone at no cost this summer, with custom dashboards and trackers to follow in the coming months. Separately, Google is expanding AI-powered Personal Intelligence to nearly 200 countries and regions in 98 languages, allowing users to connect apps like Gmail and Google Photos with Calendar support.
The Gemini app turns into an agent
Inside the Gemini app, Google introduced Spark Twinsa 24/7 personal agent that takes action on behalf of a user under their guidance. Spark runs on dedicated cloud machines, so it keeps running even when a phone or laptop is turned off, and it’s designed to check in before taking important actions. Google is rolling it out carefully, starting with trusted testers and a beta release planned for Google AI Ultra subscribers in the US next week.
A companion feature called Daily Brief runs overnight, analyzing a user’s inbox, calendar, and tasks to compile a prioritized brief, available from launch day to Google AI subscribers in the US who have connected their apps. Google also rebuilt the app’s interface with a new design language called Neural Expressive, replacing walls of text with interactive images, timelines, and embedded visuals as the response flows.
A unified platform for manufacturers
For developers, Google has expanded Antigravity in version 2.0, a desktop application that can orchestrate multiple agents in parallel, along with a command-line interface and a software development kit. Google also said it is transition the older Gemini CLI to the new Antigravity CLI, and on June 18 the legacy tool will stop serving requests for Google AI Pro and Ultra accounts and free Gemini Code Assist accounts, although enterprise access is not affected. The company released Managed Agents to the Gemini API, where a single call provides a remote Linux environment in which an agent can program, run code, and browse the web. Google AI Studio got native Android app building, and Google announced a $100 AI Ultra plan for developers, with usage limits five times higher than the Pro level, plus a Build with Gemini XPRIZE hackathon with a $2 million prize.
Shopping, work and creativity tools
presented by Google Universal Carta shopping cart that searches for deals, tracks price history, and flags product incompatibilities as soon as an item is added. It’s rolling out to Search and the Gemini app this summer, with YouTube and Gmail to follow. Across Google Workspace, the company extended AI Inbox to Gmail and added voice processing to Docs and Keep. It also launched Google Pics, an image creation and editing tool, starting with trusted testers before rolling out to Google AI Pro and Ultra subscribers worldwide this summer. Creative platform Google Flow got an agent that can design multi-step projects.
Google also unveiled Gemini for Science, a set of research tools for generating hypotheses and structuring the scientific literature, and showed that DeepMind’s Project Genie can now create interactive virtual worlds from real Street View images for subscribers to its $200-a-month plan. On the hardware front, the next milestone for Android XR is smart glasses, with the first audio glasses, made with Gentle Monster, Warby Parker and Samsung, arriving this fall. Google also detailed Ask YouTube, a conversational discovery experience now in testing, and extended SynthID and C2PA content verification to Search and Chrome.
Google I/O 2026 made it clear that Google wants Gemini to become less of a chatbot and more of a functional layer in Search, Shopping, Work, Creativity and Software Development. The strongest announcements were Gemini 3.5 Flash, Gemini Omni, Search agents, Gemini Spark, Universal Cart and the extension of Antigravity to an agent platform. The real test is availability and trust. Many of the more ambitious features are limited to subscribers, trusted testers, US betas, or summer releases, so what was released on stage was more of a model and a direction than a finished product. Google still needs to prove that agents acting on inboxes, browsers, shopping carts, and documents can be useful without becoming intrusive.
