Everything Announced at Google I/O 2026

Introduction

Google I/O 2026 arrived with a clear and deliberate statement: the era of AI as a standalone assistant is over. Furthermore, the two-day event held on May 19 and 20 at the Shoreline Amphitheatre in Mountain View, California marked the moment when Google formally declared its intention to embed Gemini not just into individual products but into the operating fabric of everything it builds. Consequently, the keynote delivered by Alphabet and Google CEO Sundar Pichai was less a product announcement and more an architectural declaration, a description of how AI agents will quietly handle tasks, monitor information, generate content, and complete transactions on behalf of users across every Google surface. This guide covers every major announcement from Google I/O 2026 in full, organized by category, so you can understand exactly what was announced and what it means for your digital life and professional workflow.

New Gemini Models

Gemini 3.5 Flash

The most consequential model launch at Google I/O 2026 is Gemini 3.5 Flash. This model surpasses Gemini 3.1 Pro on coding, agentic task execution, and multimodal benchmarks. Additionally, it runs four times faster than competing frontier models by output tokens per second, a meaningful performance advantage for applications that require real-time or near-real-time AI responses. Gemini 3.5 Flash rolled out on May 19 simultaneously across the Gemini app, Google Search AI Mode, Antigravity 2.0, and the Gemini API. A more capable Gemini 3.5 Pro model is currently in internal testing and is expected to reach developers and users within the following month.

Gemini Omni

Google DeepMind CEO Demis Hassabis presented Gemini Omni, a new model series built around the convergence of reasoning and creation. The first version, Gemini Omni Flash, accepts image, audio, video, and text as input and produces AI-generated video as output, grounded in real-world knowledge. The model also supports conversational voice editing — users can change characters, backgrounds, and scene elements by speaking naturally, without navigating any traditional editing interface. Moreover, Omni Flash is designed to roll out across the Gemini app, Google Flow, and YouTube Shorts, with developer API access arriving in a subsequent update. Future Omni versions will generate any output type from any combination of inputs.

New TPU Infrastructure: TPU 8t and TPU 8i

Supporting all of these model capabilities is a new generation of custom silicon. Google unveiled its eighth-generation Tensor Processing Units at Google I/O 2026 TPU 8t for training workloads and TPU 8i for inference. TPU 8t delivers nearly three times the raw compute performance of the previous generation while improving energy efficiency simultaneously. The TPU 8i enables Google to distribute inference workloads globally across its infrastructure. Google expects AI-related capital expenditure to reach approximately $180 billion to $190 billion this year, a figure that reflects the scale of infrastructure investment underpinning everything announced at the conference.

Gemini App: Major Updates

Neural Expressive Design Language

The Gemini app received a complete visual redesign at Google I/O 2026, built around a design language called Neural Expressive. The new aesthetic is more fluid and contextually responsive than the previous interface, and it organizes the app's growing feature set into a more navigable layout. Gemini Omni, Gemini Spark, and Daily Brief are all integrated directly into this redesigned application.

Compute-Based Usage Model

The Gemini app is transitioning from fixed daily prompt limits to a compute-based usage measurement system. Under this model, each interaction consumes resources proportional to its actual computational demands. A straightforward text query uses far less compute than a complex video generation request or an extended coding session. Limits refresh every five hours and count against a weekly compute budget rather than a fixed daily prompt allocation. Google describes this as a more equitable method of distributing platform capacity across different types of users and use cases.

Regional Dialect Support

The Gemini app now supports regional dialect options for voice responses. Users can select from a range of localized speech variations within their language, receiving AI-generated audio that reflects their regional linguistic context rather than a standardized neutral accent.

Gemini Spark: The 24/7 Personal AI Agent

Gemini Spark is the most ambitious agent product announced at Google I/O 2026. It operates continuously through dedicated cloud virtual machines, running multi-step tasks on behalf of the user even when the device is turned off, the laptop is closed, and the user is not actively engaged. Spark connects directly to Gmail, Google Sheets, Calendar, Tasks, and the broader web to coordinate complex workflows without requiring step-by-step human direction.

Spark requires user approval before taking any high-stakes actions such as sending emails on the user's behalf or completing a purchase. However, low-stakes monitoring and information gathering tasks run autonomously. A live demonstration showed Spark continuously scanning the web for updates relevant to a specific research question and delivering a summary when meaningful developments occurred.

Trusted testers received access during the week of the conference. A broader rollout to Google AI Ultra subscribers in the United States is scheduled for the following week. Access is restricted to the AI Ultra tier at launch due to the substantial compute resources required to maintain a continuous background agent for each user.

Daily Brief: Personalized Morning Digest

Daily Brief is a new Gemini feature that reads the user's Gmail, Calendar, and Tasks each morning and produces a structured daily summary. It surfaces upcoming meetings, pending tasks, and suggested next steps organized by priority based on the user's context across connected services. Daily Brief launched on May 19 and is available immediately to users on Google AI Plus, Pro, and Ultra plans in the United States.

Android Halo: Agent Status on Your Phone

Android Halo is a new on-screen interface element for Android that displays the live status of background AI agents beginning with Gemini Spark as they work. It appears as a glowing bar along the top edge of the phone screen, giving users a persistent visual signal about what their agent is currently doing without requiring them to open any application. Android Halo is coming later in 2026, starting with Gemini Spark integration.

Google Search: Redesigned for the Agentic Era

The Intelligent Search Box

Google launched what it describes as the biggest upgrade to the Search box in over 25 years. The new intelligent Search box expands dynamically as the user types, accommodating longer, more complex queries that reflect conversational intent. Additionally, the box now accepts text, images, files, videos, and Chrome browser tabs as inputs simultaneously enabling genuinely multimodal search queries from a single interface.

Unified AI Search Experience

Google merged AI Overviews and AI Mode into a single, seamless AI Search flow. Users now move fluidly from a standard search results page into an AI Overview, then into deeper AI Mode conversation all within the same interface, without switching between separate modes. This unified experience is live globally on desktop and mobile as of May 19.

Search Agents

Google I/O 2026 introduced the concept of Search agents persistent, personalized AI agents that users can create, customize, and manage directly within the Search interface. Each agent monitors a specific topic or question continuously, scanning news sources, blogs, social media, financial data, shopping listings, and sports information to surface relevant changes proactively. Search agents with Antigravity-powered app-building capabilities are coming to AI Pro and Ultra subscribers during summer 2026.

Universal Cart: Agentic Shopping

Google introduced Universal Cart, a persistent, cross-platform shopping cart powered by Gemini that follows users across Search, Gmail, YouTube, and the broader web. Universal Cart actively monitors price changes across retailers, tracks deal history, alerts users when items return to stock, and flags incompatibility issues between selected products before checkout. A demonstration showed the cart identifying that a selected PC processor and motherboard were incompatible and suggesting a replacement automatically.

The accompanying Universal Commerce Protocol enables AI agents to browse retail inventories. The Agents Payment Protocol, integrated into Gemini Spark, allows AI agents to complete purchases autonomously within spending parameters set by the user. Universal Cart launches in the United States during summer 2026, with YouTube and Gmail integrations to follow.

Ask YouTube: Conversational Video Search

Ask YouTube is a new feature inside the YouTube platform that allows users to search conversationally within video content. Rather than browsing through thumbnails or scrubbing through timelines, users ask a question and receive a structured answer that points directly to the most relevant segment within the most relevant video. Ask YouTube supports context-aware follow-up questions and is available immediately for YouTube Premium subscribers aged 18 and older in the United States through youtube.com/new.

Google Workspace: Three New Tools

Google Pics

Google Pics is a new AI image creation and editing tool built into Workspace. It treats every element within an image as a separately editable object, enabling targeted modifications without regenerating the full composition. Users can create posters, flyers, and infographics through a workflow comparable to professional graphic design tools. All output carries automatic SynthID watermarking. Pics is rolling out to trusted testers now, with broader availability for AI Pro and Ultra subscribers during summer 2026.

Docs Live

Docs Live converts natural, conversational speech including false starts, mid-sentence corrections, and casual phrasing into a finished, formatted Google Doc using Gemini. Users speak freely without concern for structure, and Gemini handles organization, formatting, and cleanup automatically. A similar capability is rolling out to Gmail and Google Keep. Docs Live is coming to AI Pro and Ultra subscribers during summer 2026.

Stitch Agent and Netlify Integration

Google's Stitch UI builder received real-time collaborative design capabilities through the new Stitch Agent at Google I/O 2026. Multiple users can now co-design simultaneously. Additionally, Stitch now exports directly to Antigravity for development handoff and supports one-click publishing to Netlify for immediate live deployment. Users can guide and reflow layouts in real time as Stitch builds based on their direction.

Google Flow and Flow Music

Google Flow, the AI creative studio for filmmakers, received Gemini Omni integration, multi-action AI agent support, and custom Flow Tools at Google I/O 2026. A dedicated mobile app for Google Flow launched alongside these updates, enabling on-the-go production work. Flow Music received a companion mobile app, targeted lyric editing within specific song sections, and the ability to generate new original compositions based on a reference segment from an existing piece.

Android XR Smart Glasses

Google I/O 2026 closed with the unveiling of Android XR smart glasses, developed in collaboration with Samsung and Qualcomm. Two versions are coming: audio glasses, which launch in fall 2026, and display glasses, which follow at a later date. Frame designs were created in partnership with Gentle Monster and Warby Parker. The glasses are compatible with both Android phones and iPhones. They deliver continuous access to Gemini through private audio output audible only to the wearer. Pricing has not yet been disclosed.

Developer Tools and Infrastructure

Antigravity 2.0 and CLI

Antigravity 2.0 is Google's updated agent-first development platform, running on Gemini 3.5 Flash. It supports parallel multi-agent workflows where complex tasks are split across specialized sub-agents that code and test simultaneously. A new command-line interface and a standalone desktop application joined the platform at Google I/O 2026, alongside standard Google Cloud privacy protections for enterprise use. The Antigravity SDK gives developers programmatic control over the agent harness for custom infrastructure deployments.

Managed Agents in the Gemini API

A new Managed Agents capability in the Gemini API removes infrastructure setup requirements for developers building agentic applications. A single API call provisions a fully operational agent with a remote sandbox. This significantly lowers the entry barrier for developers who want agent capabilities without managing the underlying orchestration infrastructure.

WebMCP

WebMCP is a proposed open web standard announced at the conference. It allows developers to expose structured tools including JavaScript functions and HTML forms so that browser-based AI agents can execute complex tasks with greater precision and reliability. The WebMCP experimental origin trial begins in Chrome 149, with Gemini in Chrome support arriving in a subsequent release.

Project Genie and Street View Integration

Project Genie, Google DeepMind's world-model platform, received a significant capability expansion at Google I/O 2026: integration with nearly twenty years of Google Street View imagery. This allows users to generate interactive, navigable 3D environments anchored in real-world geographic locations rather than purely imagined spaces. Project Genie is available globally for AI Ultra subscribers.

Pomelli

Pomelli is a new tool from Google Labs that provides AI agents for brand content creation and website design. It can generate brand books, design guidelines, and launch a complete website from a high-level description of the brand. Pomelli represents one of several new specialized agent tools expanding the Google Labs ecosystem.

SynthID and Content Credentials Expansion

SynthID Google's system for watermarking and detecting AI-generated content has now marked more than 100 billion AI-generated images and videos since its introduction. At Google I/O 2026, detection capabilities expanded to Chrome via a right-click menu on any image, to Google Search through Circle to Search on Android, and to the broader Search results interface. OpenAI, Kakao, and Eleven Labs all confirmed adoption of SynthID-compatible content tagging at the conference, expanding the system toward a cross-industry provenance standard.

AI Subscription Pricing Changes

Google restructured its AI subscription tiers significantly at Google I/O 2026. A new AI Ultra plan begins at $100 per month, substantially below the previous $250 entry point. This new tier includes five times higher Gemini app usage limits than AI Pro, 20 TB of cloud storage, priority access to Antigravity, Gemini Spark access in the United States, Project Genie globally, and a full YouTube Premium individual plan. The previous $250 plan dropped to $200 with identical capabilities. AI Pro remains at $19.99 per month and now includes YouTube Premium Lite for subscribers in select countries at no additional cost.

What Google I/O 2026 Means for Professionals

The scale and scope of Google I/O 2026 confirms that professional competency in AI tools is no longer optional. Staying ahead means understanding not just how to use these products, but how the underlying systems Gemini's model architecture, agentic frameworks, and multimodal pipelines actually work. Professionals who pursue a Tech Certification build recognized expertise across the technology domains that power the tools announced at this year's event, creating a verifiable foundation for career growth in an AI-first industry.

Furthermore, Google's Gemini ecosystem is expanding faster than any other AI platform in the consumer market. Working with Gemini effectively across Gemini 3.5 Flash, Gemini Omni, Antigravity 2.0, and the Gemini API requires structured, platform-specific expertise. Those who earn a Google Gemini Professional certification develop the technical and applied knowledge needed to build, evaluate, and deploy Gemini-powered applications across the full range of what was unveiled at this conference.

Additionally, understanding the broader landscape of AI systems including how large language models make decisions, where agentic automation introduces operational risk, and how multimodal reasoning works across text, audio, image, and video is becoming foundational for any technical role. An AI Certification provides this comprehensive grounding, equipping professionals to evaluate tools like Gemini Spark and Gemini Omni with genuine technical confidence.

Finally, the commercial implications of the announcements at Google I/O 2026 from Universal Cart and Ask YouTube to AI-powered Search agents are significant for anyone working in digital marketing and growth. A Marketing Certification equips professionals with the strategic framework to translate these new AI surfaces into campaign performance, audience targeting, and measurable commercial outcomes.

FAQs

When and where did Google I/O 2026 take place?

Google I/O 2026 was held on May 19 and 20, 2026, at the Shoreline Amphitheatre in Mountain View, California, with a simultaneous global livestream on YouTube. The consumer keynote ran approximately two hours, led by Alphabet and Google CEO Sundar Pichai.

What was the central theme of Google I/O 2026?

The central theme was the transition to agentic AI — systems that act autonomously on behalf of users across workflows rather than simply responding to individual prompts. Gemini was repositioned throughout the keynote as an always-present, always-working layer across every Google product.

What is Gemini 3.5 Flash?

Gemini 3.5 Flash is Google's newest foundation model, launched at Google I/O 2026. It outperforms Gemini 3.1 Pro on coding, agentic, and multimodal benchmarks and runs four times faster than competing frontier models on output token throughput.

What is Gemini Omni?

Gemini Omni is a new creation-focused model series that accepts any combination of image, audio, video, and text as input and outputs AI-generated video. It supports conversational voice editing and is rolling out across the Gemini app, Google Flow, and YouTube Shorts.

What is Gemini Spark?

Gemini Spark is a 24/7 personal AI agent running on dedicated cloud infrastructure. It executes multi-step tasks across Google services — including Gmail, Sheets, Calendar, and the web — continuously and autonomously, even when the user's device is inactive. It is initially available to AI Ultra subscribers in the United States.

What is the new AI Ultra pricing?

A new AI Ultra tier starts at $100 per month and includes higher usage limits, 20 TB of storage, Gemini Spark, Project Genie, and YouTube Premium. The previous $250 plan dropped to $200 with identical features. AI Pro remains at $19.99 and now includes YouTube Premium Lite.

What is Android Halo?

Android Halo is a new on-device interface element that displays the live status of background AI agents — starting with Gemini Spark — as a glowing bar along the top of the phone screen. It keeps users informed about agent activity without requiring any app to be opened.

What changed in Google Search at Google I/O 2026?

The Search box was completely redesigned for the first time in over 25 years. It now expands dynamically, accepts multimodal inputs, and connects AI Overviews and AI Mode into a single seamless flow. Persistent Search agents that monitor topics continuously are also launching for Pro and Ultra subscribers.

What is Daily Brief?

Daily Brief is a Gemini feature that reads Gmail, Calendar, and Tasks each morning and generates a personalized daily summary with prioritized tasks and suggested next steps. It launched May 19 for AI Plus, Pro, and Ultra subscribers in the United States.

What is the Universal Cart?

Universal Cart is a persistent, AI-powered shopping cart that follows users across Google Search, Gmail, YouTube, and the web. It monitors prices, tracks deals, flags product incompatibilities, and supports autonomous purchasing through the Agents Payment Protocol in Gemini Spark.

What is Ask YouTube?

Ask YouTube is a conversational search feature inside YouTube that allows users to ask questions and receive structured answers pointing directly to the most relevant video segments. It supports context-aware follow-ups and is available to US YouTube Premium subscribers aged 18 and older.

What are Android XR smart glasses?

Google, Samsung, and Qualcomm announced two versions of Android XR smart glasses at the conference. Audio glasses — launching fall 2026 — deliver private Gemini audio responses to the wearer. Display glasses follow at a later date. Both versions support Android phones and iPhones.

What is Gemini for Science?

Gemini for Science is a suite of AI research tools connecting Antigravity to over 30 major life science databases, including AlphaFold and UniProt. It includes a Co-Scientist collaborative AI research partner and experimental tools for literature analysis, hypothesis generation, and peer review assistance.

What is Antigravity 2.0?

Antigravity 2.0 is Google's upgraded, agent-first development platform. It runs on Gemini 3.5 Flash, supports parallel multi-agent workflows with simultaneous coding and testing, and adds a new command-line interface, a standalone desktop app, and enterprise-grade Google Cloud privacy protections.

What is WebMCP?

WebMCP is a proposed open web standard that lets developers expose structured tools — such as JavaScript functions and HTML forms — so browser-based AI agents can execute complex tasks more reliably. Its experimental origin trial begins in Chrome 149.

What is Project Genie's new capability at Google I/O 2026?

Project Genie now integrates with nearly twenty years of Google Street View imagery, allowing users to create navigable 3D interactive environments anchored in real-world geographic locations. It is available globally for AI Ultra subscribers.

What is Google Pics?

Google Pics is a new Workspace image creation and editing tool that treats each image element as a separately editable object. It supports creation of posters, infographics, and flyers with automatic SynthID watermarking. It rolls out to AI Pro and Ultra subscribers in summer 2026.

What is Docs Live?

Docs Live converts conversational speech including false starts and corrections into a finished, formatted Google Doc using Gemini. It is coming to AI Pro and Ultra subscribers in summer 2026, with similar features rolling out to Gmail and Google Keep.

How did SynthID expand at Google I/O 2026?

SynthID detection is now accessible through a right-click menu in Chrome, within Google Search, and via Circle to Search on Android. OpenAI, Kakao, and Eleven Labs all committed to adopting SynthID-compatible content tagging standards at the event.

What is Pomelli?

Pomelli is a new Google Labs tool that uses AI agents to help users build brand content including brand books and design guidelines and launch complete websites from high-level brand descriptions. It is part of the expanding specialized agent ecosystem announced at Google I/O 2026.