From frontier AI models and personal agents to universal shopping carts and scientific discovery tools, Google’s annual developer conference delivered a sweeping vision of what it calls the Agentic Era.
Create and Build with the Most Advanced Models
Google unveiled a series of new Gemini models and a revamped development platform designed to bring frontier intelligence to everyday tasks and long-horizon agentic work.
Gemini 3.5 Flash
- Google launched Gemini 3.5 Flash: the first in its latest series of models combining frontier intelligence with action.
- Gemini 3.5 Flash is generally available via Google’s agent-first development platform, Google Antigravity, the Gemini API, Google AI Studio, and Android Studio.
- Gemini 3.5 Flash delivers intelligence that rivals large flagship models at speeds expected from the Flash series. It outperforms Gemini 3.1 Pro on challenging coding and agentic benchmarks like Terminal-Bench 2.1 (76.2%), GDPval-AA (1656 Elo), and MCP Atlas (83.6%).
- Landing in the top-right quadrant of the Artificial Analysis index, 3.5 Flash delivers frontier-level intelligence at exceptional speed — proving that quality and low latency are no longer mutually exclusive.
- Gemini 3.5 Flash is designed for long-horizon agentic tasks. What once took a developer days or an auditor weeks, 3.5 Flash can now complete in a fraction of the time, often at less than half the cost of other frontier models. It rapidly plans, builds, and iterates to solve real-world problems, from developing new applications to maintaining codebases and preparing financial documents.
- Building on the strong multimodal foundation of Gemini 3 and 3.5 Flash, 3.5 Flash generates richer, more interactive web UIs and graphics.
- Gemini 3.5 Pro is also in development, is already in use internally, and has a rollout planned for next month.
Gemini Omni
- Gemini Omni is Google’s new model that can create anything from any input — starting with video. It combines Gemini’s intelligence with the company’s generative media models for a new level of world understanding, multimodality, and editing. Video outputs are available now, with support for additional output types to follow.
- Gemini Omni combines an intuitive understanding of physics with Gemini’s knowledge of history, science, and culture, bridging the gap from photorealism to meaningful storytelling. Its improved understanding of forces such as gravity, kinetic energy, and fluid dynamics enables more realistic scenes.
- Videos created with Omni include Google’s imperceptible SynthID digital watermark, verifiable through the Gemini app, Gemini in Chrome, and Search.
- Gemini Omni can turn any reference — image, text, video, or audio — into a single, cohesive output. Voice references are supported for audio to start, with other audio input types to follow.
Gemini Omni in the Gemini App, Google Flow, and YouTube
- Gemini Omni Flash is rolling out to all Google AI Plus, Pro, and Ultra subscribers globally through the Gemini app and Google Flow. It is also available in YouTube Shorts Remix and the YouTube Create app to users aged 18 and over at no cost.
- Creating, remixing, and editing a video is now a conversational experience in the Gemini app — enabling changes like cinematic zooms or background swaps with a simple prompt.
- Users can upload any photo or video from their camera roll, apply built-in templates with a single click, and create a custom AI avatar that looks and sounds like them.
- For creatives using Google Flow, Omni Flash allows blending of real-world inspiration with generated content and iterative conversational editing. Gemini Omni Flash also improves character consistency, meaning identity and voice are preserved across every scene.
- A new Gemini Omni upgrade in YouTube Shorts Remix lets users step directly into their favorite Shorts, adding themselves or any visual reference with a simple prompt.
Also Read: Google I/O 2026 Day 1: The Agentic Era Is Here
Search, Shop, and Discover Exactly What You’re Looking For
Google overhauled its flagship Search product, introducing agents, generative interfaces, and a universal shopping cart, describing it as its most significant Search upgrade in 25 years.
AI Search
- AI Mode, Google’s most powerful AI Search product, has surpassed 1 billion monthly users. Gemini 3.5 Flash is now the default model globally.
- AI Mode queries have more than doubled every quarter since launch. Search queries overall reached an all-time high last quarter.
- Google is launching the biggest upgrade to its Search box in over 25 years — a new, intelligent Search box reimagined with AI. Users can search using text, images, files, videos, and Chrome tabs, with AI reasoning across all inputs.
- Google is also bringing AI Overviews and AI Mode into one seamless AI Search experience. Users can flow from a question to a results page with an AI Overview, to a follow-up in AI Mode. The experience is live across desktop and mobile worldwide.
Information Agents
- Google is entering the era of Search agents, where users can create, customize, and manage multiple AI agents directly in Search. The first type — information agents — operate in the background around the clock to keep users updated on any topic, task, or project.
- Information agents search across the web and real-time data — including finance, shopping, and sports — to monitor for changes and deliver intelligent, synthesized updates that can take action.
- Information agents will be available this summer, rolling out first to Google AI Pro and Ultra subscribers.
Generative UI and Antigravity in Search
- With Google Antigravity and the agentic coding capabilities of Gemini 3.5 Flash, Search can now build custom formats for any question, on the fly.
- Custom generative UI allows Search to design layouts and assemble interactive visuals, tables, graphs, or simulations in real time. Generative UI with Antigravity is rolling out to Search this summer, free of charge.
- For ongoing projects — like planning a wedding or managing a move — Search can build entire custom experiences, such as dashboards or trackers that users can return to repeatedly.
- Custom mini-app experiences built with Antigravity will be available in Search in the coming months, starting with subscribers.
Personal Intelligence
- Google is expanding Personal Intelligence in AI Mode to nearly 200 countries and territories across 98 languages, with no subscription required.
- With AI Mode, users can securely connect apps like Gmail and Google Photos, with Google Calendar support coming soon. Users retain full control over which apps are connected.
Universal Cart
- Google is introducing Universal Cart — an intelligent shopping hub that allows users to add items while browsing Search, chatting with Gemini, watching YouTube, or reading Gmail. The cart finds deals, tracks price history, and alerts users when items come back in stock.
- The Universal Cart runs on Gemini models, improving as the models do.
- Built-in intelligent reasoning anticipates needs, flags product incompatibilities, suggests alternatives, and helps users choose between payment methods based on loyalty information and merchant offers.
- Universal Commerce Protocol (UCP) enables smooth checkout from the cart, with support for Google Pay or direct transfer to retailer sites.
- The Universal Cart is rolling out across Search and the Gemini app this summer, with YouTube and Gmail to follow.
Also Read: Why Most AI Pilots Never Make It to Production
Streamline Your Day with the Gemini App
Google introduced a suite of new personal AI features designed to run in the background, manage inboxes, and redesign the Gemini experience from the ground up.
Gemini Spark
- Gemini Spark is a 24/7 personal AI agent designed to navigate digital life, take action on a user’s behalf, and operate in the background on phones or laptops—even when devices are off.
- Spark runs on Gemini 3.5 and is built on the Google Antigravity platform. It operates autonomously and is designed to check in before taking major actions.
- Gemini Spark is early in its product journey, with safety as a top priority in this first release. A Beta is planned for Google AI Ultra subscribers in the U.S. next week.
- Future features on the Spark roadmap include texting or emailing Spark directly, creating custom sub-agents, and authorizing payments within specified budgets and merchants.
Daily Brief
- Daily Brief is a new out-of-the-box agent that organizes and prioritizes the day ahead with a personalized digest and suggested next steps.
- Gemini works overnight to analyze a user’s inbox, calendar, and tasks — connecting the dots across their digital life to surface what matters most and anticipate needs.
- Daily Brief is rolling out today to all Google AI subscribers aged 18 and over in the Gemini app, starting in the U.S. Users must have chosen to connect their Google apps.
Neural Expressive
- Google has redesigned the Gemini experience from the ground up, introducing a new design language called Neural Expressive that features fluid animations, vibrant colors, new typography, and haptic feedback.
- The tools menu has been simplified, with easier access to image, video, and music generation, as well as built-in templates.
- Responses are now laid out dynamically in real time — replacing walls of text with interactive images, timelines, and embedded visuals.
- The Gemini Live experience has been transformed — it now opens immediately and inline, using a new model that is smarter, faster, and less distracted by background noise.
- Users will soon be able to select a regional dialect. Rollout is planned for the coming weeks.
Uplevel Your Building with Agents
Google expanded its Antigravity development platform, launched Managed Agents, and announced a $2 million developer hackathon as part of a push to make agent-first development accessible to anyone.
Google Antigravity
- Google Antigravity is the company’s agent-first development platform, now massively expanding its suite of agentic capabilities, surfaces, integrations, and product features.
- Google Antigravity 2.0 is a new, standalone desktop application that serves as a central home for agent interaction, enabling users to orchestrate multiple agents executing tasks in parallel.
- Antigravity CLI is a lightweight, terminal-based product surface that lets users create new agents instantly without a graphical user interface.
- Antigravity SDK provides programmatic access to the same agent harness powering Google’s products, co-optimized for Gemini models.
- Google is also launching native voice support for Gemini audio models, along with integration across Android, Firebase, and Google AI Studio.
- The new Antigravity is agent-first by design, with new core primitives including subagents, hooks, and asynchronous task management. Gemini 3.5 Flash has been co-optimized with the Antigravity agent harness.
- Multi-day engineering efforts are collapsing into hours with a new subagent teamwork capability, available now as an early research preview in Antigravity.
- Google is unifying on Antigravity as the only platform for agent-first development. Users of Gemini CLI are encouraged to migrate to Antigravity CLI, with a published porting guide available.
- Enterprises can now connect Google Antigravity directly to their Google Cloud projects under existing enterprise terms. Existing Gemini Enterprise customers will see Antigravity rolled out in the coming months.
Google AI Studio
- A new Google AI Studio mobile app will let users capture an idea on the go and have a working prototype ready by the time they reach their desk.
- Google Workspace is now directly accessible from apps built within AI Studio, enabling dashboards on top of Sheets data, tools to organize Drive, and apps that work with team documents — without leaving AI Studio.
- Developers can now build native Android apps directly in the AI Studio build tab.
- Support for the Google Play Console in AI Studio lets developers publish apps directly to a test track, preview on an Android Emulator in the browser, and publish to Google Play’s Internal Test Track with a single click.
- First-time builders can now deploy their first two apps to Google Cloud at no cost, with no credit card required.
- Projects can now be exported directly to Google Antigravity for local development, with conversation history, project files, and secrets carrying over.
- New customization options within AI Studio include on-the-fly image generation using Nano Banana and a new edit tool that lets users annotate directly in the preview window.
Managed Agents
- Google is launching Managed Agents in the Gemini API. A single API call to the Antigravity agent provisions a remote Linux environment where the agent can reason, plan, execute code, manage files, and browse the web. Managed Agents are powered by Gemini 3.5 Flash.
- The Antigravity agent can be extended with custom instructions and skills defined in markdown files such as AGENTS.md and SKILL.md, registered as a named agent.
- Google announced the Build with Gemini XPRIZE Hackathon — a global competition with a $2 million prize pool, the largest ever for a hackathon — asking developers to use Gemini to build applications addressing pressing global challenges.
Subscriptions
- Google is launching a new $100 AI Ultra plan for developers, technical leads, knowledge workers, and advanced creators, with 5X higher usage limits and 20TB of cloud storage.
- Google AI Pro paid subscriptions now include the YouTube Premium Lite individual plan at no extra charge, adding $8.99 in monthly value with ad-free viewing on most YouTube content.
Also Read: You Gave Your AI Agent Access. Now, What Does It Have?
Get More Done at Work and Elevate Your Creativity
Google deepened its suite of Workspace AI features, introduced new creative tools, and expanded subscriber access to AI-powered drafting and task management.
AI Inbox
- AI Inbox, introduced earlier this year as a new view in Gmail, is now rolling out to all Google AI Plus and Pro subscribers in the U.S. It intelligently surfaces priority to-dos and updates.
- AI Inbox now generates personalized draft replies based on contextual information so users can review and respond in seconds.
- When a task involves reviewing a Google Doc, Sheet, or Slide, the relevant link now surfaces directly next to the to-do.
- New task management tools in AI Inbox allow users to mark individual tasks as done, dismiss unhelpful suggestions, or mark all emails in a topic as read with a single click.
- Starting this summer, Google AI Pro and Ultra subscribers will be able to query their inbox conversationally with Gmail Live.
Google Pics
- Google is introducing Google Pics, a new image creation and editing tool built on the Nano Banana model. It includes object segmentation, text editing and translation, and Workspace integrations — designed for everything from party flyers to infographics.
- Google Pics is launching today to trusted testers. It will roll out globally to Google AI Pro and Ultra subscribers and, in preview, to Workspace business customers this summer.
Docs Live and Talk to Keep
- Docs Live is a new feature that lets users create and edit documents with their voice, organizing thoughts and pulling in relevant details from Gmail, Drive, Chat and the web. Available to Google AI Pro and Ultra subscribers starting this summer.
- A new Keep feature will let users “brain dump” by voice, with Keep organizing the stream of thoughts into notes and lists automatically. Rolling out this summer to Google AI Pro and Ultra subscribers and in preview for Workspace business customers.
Google Flow
- Since its introduction at last year’s I/O, Google Flow has expanded into a full AI creative studio, launching in over 140 countries. Gemini Omni Flash is now available in Flow to Google AI subscribers globally.
- Gemini Omni Flash in Flow allows blending real-world inspiration with generated content and iterative conversational editing, resulting in improved character consistency across scenes.
- Google is introducing Google Flow Agent — enabling multi-step tasks in Flow for the first time. The agent can plan and reason through complex creative tasks under user direction.
- Built with Gemini models, Google Flow Agent can assist with brainstorming, creating multiple variations at once, batch editing assets, and organizing collections. Available to all Google Flow users globally.
- Google Flow Tools lets users vibe-code custom creative tools directly in Flow — such as video effects, hand-drawn animations, or custom shaders.
- Tools can be built with natural language, no coding experience required, and shared with other Flow users for remixing. All Flow users can access existing tools; Google AI subscribers can create and remix them.
Google Flow Music, Pomelli, and Stitch
- Google Flow Music uses Gemini Omni to let users direct shareable music videos conversationally. New refinement capabilities allow editing specific sections of a song — changing lyrics, genre, instruments, and more.
- Pomelli is adding new ways to build brand content and design websites.
- Stitch now offers real-time design and steering through text or voice, with support for importing existing codebase and design files to keep builds on-brand.
Drive Scientific Breakthroughs and Discoveries
Google introduced Gemini for Science, a collection of experimental tools designed to accelerate research at every stage — from literature review to hypothesis generation and computational discovery.
- Google introduced Gemini for Science, a new collection of science tools and experiments designed to expand the scale and precision of scientific exploration. Three new experimental tools are available on Google Labs.
- Hypothesis Generation, built with Co-Scientist, simulates the scientific method. It uses a multi-agent idea tournament to generate, debate and evaluate hypotheses, with claims verified and supported by clickable citations.
- Computational Discovery, built with AlphaEvolve and Empirical Research Assistance (ERA), generates and scores thousands of code variations in parallel — allowing scientists to test novel modeling approaches in fields like solar forecasting or epidemiology in a fraction of the time previously required.
- Literature Insights, built with NotebookLM, searches scientific literature and structures results into tables with custom, searchable attributes. Researchers can use chat to uncover nuances and create reports, slide decks, infographics, and audio and video overviews.
- Access to all three experimental tools opened gradually from May 19. Users can visit labs.google/science to register interest.
- As part of Gemini for Science, Google is launching Science Skills — a specialized bundle integrating insights from over 30 major life science databases including UniProt, AlphaFold Database, AlphaGenome API and InterPro. Available from May 19 on GitHub and for all Google Antigravity users.
- Google has created dedicated pilots with leading scientific conferences including ICML, STOC and NeurIPS to develop tools for agentic peer review and scientific validation, including the experimental Paper Assistant Tool (PAT) and ScholarPeer.
Also Read: Healthcare Is Scaling AI Without the Infrastructure to Manage It
Transform the Way You Learn and Explore
Google introduced new conversational search capabilities for YouTube and announced the next milestone for Android XR, including intelligent eyewear made with major fashion brands.
Ask YouTube
- Ask YouTube is a new conversational search experience on YouTube. Users can ask complex queries — such as how to teach a child to ride a bike — and receive structured, interactive responses compiled from across YouTube’s entire catalog, including long-form videos and Shorts.
- Ask YouTube is beginning to roll out this month on desktop as an experiment to a subset of users searching in English in the U.S.
Android XR
- The next milestone for Android XR is intelligent eyewear, available in two types: audio glasses that offer spoken assistance, and display glasses that show information as needed.
- Google’s first audio glasses, made in partnership with Gentle Monster, Warby Parker and Samsung, will arrive this fall and will be compatible with Android and iOS devices.
SynthID
- Google launched SynthID three years ago as a watermarking technology that embeds imperceptible signals into AI-generated content. Verification has already been used 50 million times globally. The capability is now expanding to Search and to Chrome in the coming weeks.
- Users can check whether an image was made with AI using Search features like Lens, AI Mode, and Circle to Search, as well as Gemini in Chrome.
- Google is also adding verification for C2PA Content Credentials, allowing users to check whether content is an unaltered original or has been modified, and by what tools. Rolling out in the Gemini app now, with Search and Chrome to follow.
- OpenAI, Kakao and ElevenLabs are among the companies bringing SynthID technology to their own AI-generated content, extending the watermarking standard across the web.
