Every Major AI Announcement at Google I/O 2026
Every Major AI Announcement at Google I/O 2026 (The Complete Breakdown)
Category: Google, AI News, Google I/O 2026
Quick Answer: At Google I/O 2026, held May 19 to 20 at Shoreline Amphitheatre in Mountain View, California, Google announced two new AI model families (Gemini 3.5 and Gemini Omni), a 24/7 autonomous personal agent called Gemini Spark, the biggest Search redesign in 25 years with Information Agents, a Gemini-powered Universal Cart for AI shopping, a new design app called Google Pics, AI voice editing in Docs and Gmail (Docs Live and Gmail Live), Android XR smart glasses in partnership with Samsung and Qualcomm, Ask YouTube, Daily Brief, Antigravity 2.0 for developers, and SynthID watermarking now covering over 100 billion AI-generated images and videos. The central theme of the entire event was the shift from AI assistants that answer questions to AI agents that take action.
If last year's Google I/O was about showing that Gemini could do everything, this year's was about proving it actually does.
Sundar Pichai opened the keynote on May 19 by saying they are now "firmly in the agentic Gemini era," and everything that followed made that statement hard to argue with. Two new model families. A 24/7 autonomous agent running on Google Cloud. The largest Search redesign in a generation. Smart glasses that actually work. Over 85 sessions, codelabs, and developer tools spanning two days.
This is not a summary of what Google hopes to do. This is a breakdown of what Google announced, what is available right now, and what it actually means for how you use the internet, your phone, and your work every day.
Here is every major AI announcement from Google I/O 2026, organized by what actually matters.
1. Gemini 3.5 Flash and Gemini 3.5 Pro: The New Foundation Models
The most important technical announcement at I/O 2026 is Gemini 3.5 Flash, and it is available right now.
Gemini 3.5 Flash combines frontier intelligence with the ability to perform agentic tasks. It surpasses 3.1 Pro in coding, agentic, and multimodal benchmarks, with the cost and speed of the Flash series at 4x faster than other frontier models in terms of output tokens per second.
That last number is worth sitting with. Four times faster than other frontier models in output tokens per second. For anyone building on top of Gemini through the API, or using Antigravity for development, this is a meaningful performance jump at the same or lower cost point as 3.1.
Gemini 3.5 Flash is rolling out starting today in the Gemini app, Search, Antigravity 2.0, and Gemini API. Gemini 3.5 Pro is currently in testing and will be available next month.
Gemini 3.5 Pro, expected next month, is positioned as the premium tier. Early access users report that it significantly outperforms 3.1 Pro on complex reasoning and coding tasks, though Google has not yet released full benchmark comparisons for the Pro variant.
What this means for you: If you use the Gemini app or build on the Gemini API, you are already running on a faster, more capable model as of today. The coding and agentic improvements are particularly relevant for anyone using Antigravity.
2. Gemini Omni: A New AI That Turns Anything into Video
Gemini Omni is the most visually striking announcement from the keynote, and DeepMind CEO Demis Hassabis took the stage personally to introduce it.
Gemini Omni can create anything from any input, starting with video, and is a leap forward in world understanding, multimodality and editing.
Gemini Omni is a new series of models that combines Gemini's reasoning capabilities with creation. Gemini Omni Flash today accepts image, audio, video, and text input and outputs video grounded in real-world knowledge and that can be easily edited.
The demos on stage showed Gemini Omni simulating complex physical concepts like kinetic energy and gravity, translating scientific ideas into short digestible videos, and generating cinematic video from a reference image. Perhaps most impressively, it can take a short clip from your phone and transform it with a conversational prompt — applying a cinematic zoom, changing the background, or shifting the mood — without any video editing knowledge required.
Gemini Omni turns any reference including image, text, video or audio into a single, cohesive output. Gemini Omni Flash is rolling out now to all Google AI Plus, Pro and Ultra subscribers globally through the Gemini app and Google Flow. It is also available today in YouTube Shorts Remix and the YouTube Create app to users aged 18 and over at no cost.
What this means for you: Content creators, YouTube creators, and marketers now have a free-to-access video generation tool built directly into YouTube Shorts that takes a reference input and produces a full video. This is not a separate subscription. It is built into tools billions of people already use.
3. Gemini Spark: Your First 24/7 AI Agent
This is the announcement that will generate the most conversation over the next few months, and for good reason.
Pichai's biggest announcement was Gemini Spark, a 24/7 personal AI agent that runs on Google Cloud and keeps working even when your phone is locked or your laptop is closed.
The capabilities demonstrated on stage are a significant step beyond anything previously available in a consumer AI product:
- Parsing your credit card statements for hidden subscriptions
- Monitoring your inbox and flagging deadlines from school emails
- Writing meeting notes into a Google Doc and emailing it out
- Drafting project kickoff emails from a quick voice note
- Planning and executing multi-step tasks while you sleep
Google described Spark as a system that can run continuously on dedicated virtual machines within Google Cloud infrastructure.
Gemini Spark integrates with Gmail, Docs, and other Google Workspace apps before expanding to other third-party tools via MCP over the summer. Gemini Spark will be available next week to Google AI Ultra subscribers in the US.
The MCP (Model Context Protocol) integration coming this summer is significant. It means Spark will be able to connect to third-party apps through a standard protocol, the same one used by Anthropic's Claude and other leading AI tools. That opens Spark to a wide ecosystem of integrations without Google having to build each one directly.
Spark will be accessible in the Gemini app, but Google also plans to bring it to other applications like Gmail and Google Workspace, plus operating systems like Android and iOS. Spark will roll out first in beta to Google AI Ultra subscribers in the US next week.
Google AI Ultra is priced at $100 per month. Broader rollout to lower tiers is expected later in 2026.
What this means for you: Gemini Spark is the most ambitious consumer AI agent announced by any major company in 2026. It competes directly with Perplexity's Personal Computer feature and OpenAI's Operator. The key differentiator is Google's native access to Gmail, Calendar, Docs, and Workspace data, which gives Spark a context advantage that third-party agents cannot easily replicate.
4. Daily Brief: Your AI-Powered Morning Digest
Alongside Spark, Google announced a lighter-weight agent that starts working immediately for all paid subscribers.
A new Gemini agent called Daily Brief creates a personalised morning digest from your Gmail, Calendar, and Tasks. It is rolling out to AI Plus, Pro, and Ultra subscribers in the US starting today.
Daily Brief is a personalized digest of the day ahead. Sifting through your Gmail, Calendar, and Tasks, it prioritizes and organizes what you need to do, while suggesting next steps.
Daily Brief is available now for Google AI Plus ($20/month) and above. It is the most immediately accessible agentic feature from the entire keynote for most users. Think of it as a smarter version of a morning briefing that actually knows what is in your inbox and on your calendar, rather than giving you generic news headlines.
5. The Biggest Search Redesign in 25 Years
Google described its Search redesign as the biggest upgrade to Search in over 25 years, and the features announced back that claim up.
The core shift is from Search as a place you visit to get results, to Search as an ongoing system that works for you continuously. Three features define this new model.
Information Agents in Search: Google announced information agents in Search, AI systems that users can set up to continuously monitor topics and return useful updates or actions in the background. These are persistent agents that you configure once. You tell Search to monitor a topic, a price, a news story, or a regulatory change, and it surfaces updates to you proactively rather than waiting for you to search again.
Generative UI in Search: Search will be able to build a custom response on the fly with dynamic layouts, interactive widgets, and more for queries. It uses Antigravity and 3.5 Flash. Search can create tools, trackers, widgets, and dashboards. Generative UI in Search is rolling out this summer for everyone with no charge.
Conversational Search: Pichai said Search now feels more like an ongoing conversation, with users asking longer and more complex questions.
The redesigned Search box is itself the most visual signal of this shift. Google called it "the biggest upgrade to Search in over 25 years," which is a significant claim for a product that has defined how a generation uses the internet.
6. Ask YouTube: AI Navigation Across Video Content
One of the new AI-powered experiences tied to the Search vision is Ask YouTube, designed to help users navigate videos more efficiently by surfacing relevant clips and jumping directly to the most useful sections. Google said the feature is currently being tested and will launch more broadly in the US this summer.
Ask YouTube effectively turns the world's largest video library into a searchable knowledge base. Instead of watching a 45-minute tutorial to find the three minutes that answer your question, Ask YouTube finds the exact clip and takes you directly there. This is meaningful for education, how-to content, and any long-form video category where most of the value is concentrated in a fraction of the runtime.
7. Universal Cart: AI Shopping Across the Entire Internet
This announcement has significant implications for e-commerce and is one of the most practically impactful features announced at I/O 2026.
Universal Cart is a Gemini-powered shopping cart and agentic hub for shopping in the Gemini app, YouTube, and Gmail. Once you add a product to your cart, Google finds deals, price drops, price history insights, and stock alerts.
The demo on stage showed a scenario that captures how different this is from current shopping: you are building your first custom PC and add a few parts from several retailers to your cart. The Universal Cart proactively flags any product incompatibilities and suggests alternatives. It also understands your payment method perks, loyalty information and merchant offers to help you choose.
The Universal Commerce Protocol allows AI agents to browse the world's biggest inventories and handle the entire purchase, right from Search. You can look for a product on Google, and it will find the lowest price on any of the commerce partners' websites, then allow you to check out through Search.
Universal Cart is coming to Search and the Gemini app in the US this summer, with YouTube and Gmail to follow.
What this means for you: Universal Cart is Google's most direct move into the commerce layer of the internet. It combines what Amazon does (product discovery and purchase) with what Google does (search and information) and adds an AI agent layer on top. For anyone who shops online, this is a meaningful change to how product research and purchase will work.
8. Docs Live, Gmail Live, and Voice-First Productivity
Google Workspace got a significant AI upgrade at I/O, anchored by a new voice-first editing paradigm.
Docs Live lets you edit, format, and restructure documents entirely with your voice. The demo showed someone telling Docs to move a section to the top, make it bold, and add a summary paragraph — all without touching the keyboard. Being able to tell Docs to move sections up top and format them with bold or italics instead of having to do so with your fingers could be much easier, with particular accessibility implications.
Gmail Live, coming this summer, lets you conversationally search your emails. Starting this summer, Google AI Pro and Ultra subscribers will be able to talk to their inbox with Gmail Live, asking specific queries without having to dig through threads.
AI Inbox got updates too. Starting now, in AI Inbox, if a task requires reviewing a Google Doc, Sheet or Slide, the relevant link surfaces right next to your to-do. Starting this summer, Google AI Pro and Ultra subscribers will be able to talk to their inbox with Gmail Live, asking specific queries without having to dig through threads.
For professionals who spend large portions of their working day in Gmail and Docs, these features represent the most meaningful change to everyday productivity tools since the introduction of keyboard shortcuts.
9. Google Pics: A New AI Design App for Workspace
Google announced new voice capabilities in Gmail, Docs and Keep, a new design tool called Google Pics, and updates to AI Inbox.
Google Pics is a new AI-powered design and image generation app built directly into Google Workspace, announced at I/O 2026. It positions Google as a direct competitor to Canva in the AI-assisted design space.
You can use Pics to upload existing images to Workspace and edit them. It is possible to remove and resize both foreground and background elements. Each generation is fingerprinted with Google's SynthID watermark.
The SynthID fingerprinting is an important detail. Every image generated through Google Pics is automatically watermarked with an invisible identifier, making it possible to verify whether an image was AI-generated even after editing or compression. This is part of Google's broader push on AI content authenticity.
10. Antigravity 2.0: Google's Agent-First Developer Platform
For developers, Antigravity 2.0 is the most important announcement at I/O 2026. It is Google's answer to Cursor and Claude Code.
Antigravity 2.0 is Google's coding tool, the equivalent of Copilot, Codex, and Claude Code. Gemini 3.5 Flash is 12x faster in Antigravity, which optimizes token use. Antigravity 2.0 is available globally for everyone.
Antigravity is Google's tool for developers to create AI agents and programs at scale. It will now use Gemini 3.5 Flash and allow for faster development cycles. Google announced at I/O that Antigravity will get new Google Cloud standard privacy protections, a standalone desktop app dubbed Antigravity 2.0, and a new command-line interface for developers who prefer staying in Terminal.
The developer keynote added more detail: Instead of writing complex orchestration code, you can define everything in markdown files like AGENTS.md and SKILL.md and register them as a named agent.
Google AI Studio now includes native Kotlin support to vibe code Android apps. With Google Workspace integrations and a one-click deploy to Cloud Run along with support for Firebase services, you can now build and launch full-stack apps directly within AI Studio. And if you want to continue building, you can also seamlessly export your complete project state to Google Antigravity.
What this means for developers: Antigravity 2.0 with Gemini 3.5 Flash closes a meaningful gap with Cursor on speed and cost efficiency. The markdown-file-based agent definition is particularly interesting — it borrows the AGENTS.md convention from the broader AI ecosystem and makes it a first-class concept in Google's developer platform.
11. Android XR Smart Glasses: Real Hardware, Real Partnerships
Smart glasses have been promised for years. At I/O 2026, Google showed them working.
Google, Samsung, and Qualcomm unveiled the first Android XR smart glasses, featuring designs from Gentle Monster and Warby Parker. Two versions are coming: audio glasses launching fall 2026 and display glasses to follow later. The glasses work with both Android phones and iPhones.
The capabilities shown on stage include: directions through the glasses without taking out your phone, sending texts and taking photos by voice, real-time language translation overlaid in your field of view, and Gemini integration for in-context assistance. The Gentle Monster and Warby Parker design partnerships are significant because they signal that Google is learning from the failure of Google Glass — fashion matters as much as function for wearable adoption.
More details are expected at Samsung Galaxy Unpacked in July 2026. Pricing has not yet been disclosed.
Android Halo was also announced for phones: Android Halo is a new on-screen indicator that shows you the live status of AI agents like Gemini Spark as they work in the background. It appears as a glowing bar at the top of your phone screen. Coming later this year, starting with Gemini Spark.
12. SynthID and C2PA: Google's Answer to Deepfakes
One of the less flashy but most important announcements at I/O 2026 is the significant expansion of SynthID.
Google emphasized ethical AI development with SynthID watermarking for over 100 billion images and videos.
Google is expanding digital watermarking with SynthID and C2PA Content Credentials to help you identify AI-generated media.
Google is rolling out SynthID and C2PA verification to Search and Chrome, so users can more easily identify AI-generated imagery.
C2PA (Coalition for Content Provenance and Authenticity) is an industry standard that embeds verifiable metadata into digital content. By rolling SynthID and C2PA out to Search and Chrome, Google is working toward a future where every AI-generated image you encounter on the web carries a verifiable tag identifying it as AI-generated, even if it has been edited, compressed, or shared multiple times.
C2PA content credentials are coming to Gemini and Chrome.
13. Gemini for Science: AI Research Tools
Gemini for Science is a collection of science tools, and there is also a Co-Scientist collaborative AI research partner.
Co-Scientist is a research assistant specifically designed to work alongside professional scientists on complex, multi-step research tasks. Combined with Google's announcement that AI-related publications in the natural, physical, and life sciences increased 26% to 28% year over year, this signals Google's serious intent to become a foundational tool in scientific research workflows.
14. TPU 8: The Infrastructure Behind Everything
A decade ago, Google announced its very first commercial tensor processing unit, or TPU, on the I/O stage. Google recently announced its 8th generation of TPUs at Cloud Next. For the first time, Google has taken a dual chip approach with specialized architectures for training and inference: TPU 8t and TPU 8i. TPU 8t is optimized for large-scale pretraining, and it is nearly three times the raw computing power of the previous generation.
The TPU 8i is optimized for inference, meaning the cost and speed of running Gemini models for everyday users should improve significantly as this infrastructure rolls out. The 12x speed improvement in Antigravity cited earlier is partly attributable to these infrastructure advances.
The creation of the world's largest AI training cluster is designed to accelerate model training and improve computational efficiency.
15. Pricing Updates: Google AI Plans Restructured
Google shook up its subscription offerings at I/O 2026, and the AI Pro tier is coming out on top. The $19.99 per month plan now includes YouTube Premium Lite at no additional charge.
The Gemini app is moving from daily prompt limits to a compute-used model that factors the complexity of your prompt, features you use, and length of your chat.
The shift from prompt-count limits to compute-based limits is a meaningful change for power users. Simple conversational queries will use less of your monthly compute budget than complex agentic tasks, meaning casual users get effectively unlimited access while heavy users are billed proportionally to what they actually consume.
Google AI Ultra includes 5X higher usage limits in the Gemini app and Antigravity than the AI Pro plan, plus 20TB of cloud storage and more features to help accelerate development cycles. Google AI Ultra is priced at $100 per month.
Everything Announced at Google I/O 2026: Quick Reference Table
| Announcement | Category | Available |
|---|---|---|
| Gemini 3.5 Flash | AI Model | Now |
| Gemini 3.5 Pro | AI Model | Next month |
| Gemini Omni Flash | AI Model / Video | Now |
| Gemini Spark | AI Agent | Next week (Ultra) |
| Daily Brief | AI Agent | Now (Plus+) |
| Information Agents in Search | Search AI | Summer 2026 |
| Ask YouTube | YouTube AI | Summer 2026 US |
| Universal Cart | Shopping AI | Summer 2026 US |
| Docs Live | Productivity | Now |
| Gmail Live | Productivity | Summer 2026 |
| Google Pics | Design AI | Now |
| Antigravity 2.0 | Developer | Now |
| Android XR Audio Glasses | Hardware | Fall 2026 |
| Android XR Display Glasses | Hardware | TBD |
| Android Halo | Mobile | Later 2026 |
| SynthID + C2PA in Search/Chrome | Safety | Rolling out |
| Gemini for Science | Research | Now |
| TPU 8t and 8i | Infrastructure | Rolling out |
What Google I/O 2026 Actually Means
The pattern running through every announcement at I/O 2026 is the same: Google is collapsing the distance between intent and action.
Search used to be a place you went to find information. Now it creates interactive tools in response to your query. Gmail used to be a place you read and wrote email. Now an agent reads it for you and acts on it before you open the app. Shopping used to require visiting multiple stores. Now Universal Cart monitors the entire internet and executes purchases on your behalf.
The Gemini model is the AI layer underneath all of it. Gemini is becoming the AI layer beneath Search, Workspace, Android, coding, and more. That is not promotional language. It is an accurate description of what was announced.
The two things worth watching closely after I/O: how well Gemini Spark performs in real-world use (the demos were impressive but demos are always impressive), and how the Universal Cart changes e-commerce once it launches this summer. Those two features, more than any model benchmark, will determine whether Google I/O 2026 is remembered as the event where Google declared its agentic era, or the event where it actually delivered it.
Frequently Asked Questions
What was announced at Google I/O 2026?
Google I/O 2026, held May 19 to 20, announced Gemini 3.5 Flash and Gemini Omni as new AI models, Gemini Spark as a 24/7 autonomous AI agent, Daily Brief for morning personalized digests, the biggest Search redesign in 25 years with Information Agents, Ask YouTube, Universal Cart for AI-powered shopping, Docs Live and Gmail Live for voice editing, Google Pics for AI image design, Antigravity 2.0 for developers, Android XR smart glasses in partnership with Samsung, and SynthID watermarking now covering over 100 billion AI images and videos.
What is Gemini Spark?
Gemini Spark is a 24/7 personal AI agent announced at Google I/O 2026. It runs on dedicated virtual machines in Google Cloud, meaning it keeps working even when your phone is locked or your laptop is closed. It can parse emails, monitor your calendar, draft documents, send communications, and complete multi-step tasks on your behalf. It integrates with Gmail, Docs, and Google Workspace at launch, with third-party app support via MCP coming summer 2026. Spark rolls out first to Google AI Ultra ($100/month) subscribers in the US next week.
What is Gemini Omni?
Gemini Omni is Google's new AI model series announced at Google I/O 2026. It combines Gemini's reasoning capabilities with video creation, accepting image, audio, video, and text as input and outputting video. Gemini Omni Flash is available now to all Google AI Plus, Pro, and Ultra subscribers in the Gemini app, and is free in YouTube Shorts Remix for users aged 18 and over.
What is Gemini 3.5 Flash?
Gemini 3.5 Flash is the newest model in the Gemini 3.5 series, released at Google I/O 2026. It surpasses Gemini 3.1 Pro on coding, agentic, and multimodal benchmarks while running at 4x faster output token speed than other frontier models. It is available now in the Gemini app, Search, Antigravity 2.0, and via the Gemini API. Gemini 3.5 Pro is in testing and expected next month.
What is Google Universal Cart?
Universal Cart is a Gemini-powered shopping system announced at Google I/O 2026. It creates a single intelligent shopping cart that works across Google Search, the Gemini app, YouTube, and Gmail. Once you add a product, Universal Cart monitors price drops, alerts you to stock changes, flags product incompatibilities, and finds deals across merchants. It supports Google Wallet, loyalty cards, and merchant offers. It is coming to Search and the Gemini app in the US in summer 2026.
What are Android XR smart glasses?
Android XR smart glasses are wearable AI glasses announced at Google I/O 2026 in partnership with Samsung and Qualcomm, featuring designs from Gentle Monster and Warby Parker. Two versions are planned: audio glasses launching fall 2026 and display glasses to follow. The glasses work with both Android phones and iPhones, and support Gemini for directions, messaging, photography, and real-time language translation. Pricing has not been disclosed.
What is Antigravity 2.0?
Antigravity 2.0 is Google's updated agent-first developer platform, announced at Google I/O 2026. It uses Gemini 3.5 Flash and is 12x faster than the previous version for token processing. It includes a new standalone desktop app, a command-line interface, and the ability to define AI agents using markdown files (AGENTS.md and SKILL.md). It is Google's direct competitor to Cursor, Claude Code, and GitHub Copilot. Antigravity 2.0 is available globally for all developers now.
What is Google Docs Live?
Docs Live is a new voice-editing feature in Google Docs announced at Google I/O 2026. It allows you to edit, format, move, and restructure content in a Google Doc entirely by voice, without using the keyboard. Versions for Gmail and Google Keep are also coming. It is available now and has significant accessibility implications for users with motor difficulties.
Final Thoughts: The Agentic Era Has a Start Date
Google I/O has always been about showing where the company's products are heading. But I/O 2026 is different from recent years in one specific way: most of what was announced is available today, or within weeks.
Gemini 3.5 Flash is live. Gemini Omni Flash is live. Antigravity 2.0 is live. Daily Brief is live for paid subscribers. Docs Live is live.
Gemini Spark, Universal Cart, Ask YouTube, Gmail Live, and the Android XR glasses are the big features still coming over the next few months. But the foundation they are built on is already in your hands.
Sundar Pichai said at the start of the keynote that the AI cycle has moved to a phase where people want to see the value in the products they use every day. For the first time at a Google I/O in years, the announcement list delivers on that expectation. The agentic era did not just get declared at I/O 2026. It started shipping.
All information in this article is sourced directly from Google's official I/O 2026 blog, developer keynote, and announcements published May 19 to 20, 2026.
