Google I/O 2026 wrapped up its opening keynote on May 19, 2026, at the Shoreline Amphitheatre in Mountain View, California — and once again, AI stole the show. A brand new Gemini 3.5 model family, a “24/7” personal AI agent called Gemini Spark, a complete overhaul of Google Search, and the long-awaited Samsung Intelligent Eyewear: this year’s developer conference made one thing clear, Google is going all-in on agentic AI.
If you missed the nearly two-hour livestream, don’t worry. Here’s our complete breakdown of every major announcement from the Google I/O 2026 keynote — what it means, when you’ll get it, and why it matters.
Table of Contents
When and Where Was Google I/O 2026?
Google I/O 2026 took place on May 19–20, 2026 at the Shoreline Amphitheatre in Mountain View, California, with the main keynote starting at 10 AM PT and delivered by Alphabet and Google CEO Sundar Pichai. As in previous years, all keynotes, developer sessions, and demos were live-streamed for free on Google’s official YouTube channel and on io.google.
This year’s event spanned two full days, with the consumer-facing keynote on Day 1 and a deeper developer keynote following soon after. Over 85+ sessions, codelabs, and workshops are available on demand starting May 21.
1. Gemini 3.5 Flash and Gemini 3.5 Pro: A New AI Model Family
The biggest announcement of Google I/O 2026 was the debut of the Gemini 3.5 model family, which Google describes as “a major leap forward in building more capable, intelligent agents.”
Gemini 3.5 Flash highlights:
- 4x faster than other frontier AI platforms and models in output tokens per second
- Outperforms the previous Gemini 3.1 Pro on coding, agentic, and multimodal benchmarks
- Designed for long-horizon agentic tasks and can deploy teams of subagents
- Rolling out immediately across the Gemini app, AI Mode in Search, Antigravity 2.0, and the Gemini API
- Becomes the default model across several Google services
Gemini 3.5 Pro — the heavier, more capable variant — is currently in internal testing and limited rollout, and is expected to be widely available in June 2026.
For developers and businesses, Gemini 3.5 Flash is positioned as the new sweet spot between speed, cost, and intelligence — ideal for building apps, updating codebases, drafting financial documents, and powering autonomous workflows.
2. Gemini Spark: Google’s New “24/7 Personal AI Agent”
Possibly the most futuristic announcement of the keynote, Gemini Spark is a new personal AI agent that lives inside the Gemini app — but doesn’t need your device to be on to keep working.
Here’s what makes Gemini Spark different:
- It runs on virtual machines in Google Cloud, so it can continue tasks 24/7, even when your phone is asleep
- Powered by Gemini 3.5
- Integrates natively with Gmail, Google Docs, Slides, and Sheets, plus third-party apps like Canva, OpenTable, and Instacart
- Acts as an “active partner that does real work on your behalf and under your direction” — not just answers questions
Gemini Spark is rolling out in beta to trusted testers and Google AI Ultra subscribers in the US, starting the week after I/O.
3. Google AI Ultra: The New $100/Month Premium Plan
Google quietly reshuffled its AI pricing landscape with the launch of a new Google AI Ultra plan priced at $100 per month. The tier is aimed squarely at:
- Developers
- Content creators
- AI power users
- Enterprise prosumers
AI Ultra subscribers get early access to Gemini Spark, higher rate limits, and priority access to new Gemini features as they roll out. The price puts Google in direct competition with OpenAI’s ChatGPT Pro and Anthropic’s Claude Max tiers.
4. Gemini Omni: The New World Model for AI Video
Google introduced Gemini Omni, a brand-new multimodal “world model” that can generate polished video content from virtually any input — text, images, audio, video, or any combination. We’ll test it out in the coming days of how good it actually is in comparison of competitors.
Unlike its predecessor Veo 3, which was primarily a text-to-video model, Gemini Omni:
- Accepts mixed inputs (text + image + audio + video)
- Produces videos “grounded in Gemini’s real-world knowledge”
- Allows users to edit videos through natural conversation
- Is being used internally to create visual explainers for complex topics
Google also confirmed it is expanding its SynthID watermarking technology to clearly label AI-generated content created with Gemini Omni — an attempt to address growing concerns about AI misuse and deepfakes.
5. Google Search Gets Its Biggest Upgrade in Nearly 30 Years
Google declared that Search is getting its biggest makeover in almost three decades. The new experience is now powered end-to-end by Gemini 3.5 Flash and introduces several key features:
- Intelligent Search box that expands as you type, reflecting the trend of longer, more conversational queries
- AI-powered query suggestions that anticipate what you mean — going beyond traditional autocomplete
- AI Mode upgraded to handle multi-step reasoning and agentic tasks
- Deeper integration with Gmail, Maps, YouTube, and Shopping
This is Google’s clearest signal yet that traditional “10 blue links” search is being replaced by an AI-first, answer-first experience.
6. Universal Cart: AI-Powered Shopping Across the Web
Google unveiled the Universal Cart, which the company calls “a truly intelligent shopping cart” and a new agentic hub for shopping across Google.
Key features:
- Works across merchants and across services — you can add items while browsing Search, chatting with Gemini, watching YouTube, or reading Gmail
- Proactively tracks deals, price drops, and restocks using Gemini AI
- Helps prevent buying mistakes (for example, recognizing incompatible PC parts in a build)
Behind the scenes, Google also launched the Universal Commerce Protocol (UCP) — an open standard for unifying digital commerce — with partners including Amazon, Shopify, and Walmart.
7. Ask YouTube: A Conversational Way to Find Videos
Google “entirely reimagined” how users find videos with a new feature called Ask YouTube. Instead of typing a keyword and scrolling through results, users can now have a conversation with YouTube to find the exact video — or specific moment within a video — they’re looking for.
The feature is powered by Gemini 3.5 and combines YouTube’s massive video library with Gemini’s reasoning capabilities to surface results based on intent, context, and natural language.
8. Samsung Intelligent Eyewear: Google’s Answer to Meta Ray-Bans
The keynote closed with one of the most anticipated reveals: Samsung Intelligent Eyewear, developed in partnership with Google, Warby Parker, and Gentle Monster.
What you need to know:
- Two design lines: a Warby Parker version (refined, timeless eyewear) and a Gentle Monster version (bold, fashion-forward sunglasses)
- Built on Android XR and powered by Gemini
- Acts as a companion device to your smartphone — voice-controlled, hands-free
- Deep integrations with Google Maps (for turn-by-turn directions in your vision) and the Gemini app
- Launching Fall 2026 as part of each eyewear brand’s full collection
Samsung positions this as the next major form factor in its Galaxy ecosystem — a direct shot at Meta’s Ray-Ban smart glasses lineup. Pricing has not yet been confirmed.
Google also confirmed that Xreal’s Aura smart glasses running Android XR will ship before the end of 2026, expanding the platform’s hardware lineup.
9. Antigravity 2.0: Google’s Agent-First Developer Platform
For developers and vibe coding, Google unveiled Antigravity 2.0 — a significant upgrade to its agent-first development platform — along with a brand-new Antigravity CLI.
Key capabilities include:
- Spinning up specialized subagents to tackle complex workflows
- Cross-platform terminal sandboxing
- Credential masking and hardened Git policies for security
- New tools for building modern Android and web apps using AI agents
This is Google’s most serious push yet into the developer-focused AI coding space, putting it in direct competition with offerings from Cursor, GitHub Copilot, and Anthropic’s Claude Code.
10. Gemini for Science and Verify AI
A few more announcements worth noting:
- Gemini for Science: A new initiative offering AI tools to accelerate research and improve scientific predictions, including hurricane forecasting
- Gemini Verify AI: A new tool — backed by Nvidia and OpenAI — designed to help users verify whether images and content are AI-generated or authentic
- Expanded SynthID watermarking across more Google AI products
- A sneak peek at Android 17 (codenamed Cinnamon Bun), with a stable release planned for June 2026
- Updates across Android Auto, Wear OS, Android TV, and Google Workspace
Google I/O 2026 Key Takeaways
If Google I/O 2025 was about positioning Gemini as a universal assistant, Google I/O 2026 was about turning it into an autonomous agent that takes action on your behalf. The keynote’s biggest themes were:
- Agentic AI is the new normal — Gemini Spark and Antigravity 2.0 show Google moving from “AI that helps” to “AI that does”
- Search is being rebuilt around AI — the traditional search box is on its way out
- Smart glasses are finally real — with Samsung, Warby Parker, and Gentle Monster behind them, Android XR has a credible shot at the form factor
- Premium AI is the new battleground — the $100/month AI Ultra plan signals Google sees serious money in power users
- Open ecosystems matter — UCP, partnerships with Amazon and Shopify, and the expanding Android XR lineup show Google playing the long game
Final Thoughts
Google I/O 2026 was less about flashy hardware and more about quietly reshaping the entire AI software stack — from how you search, to how you shop, to how your digital assistant operates while you sleep. With Gemini 3.5 setting a new performance bar and Gemini Spark introducing always-on agentic AI, Google has clearly drawn its line in the sand against OpenAI, Anthropic, and Meta.
The real test now? Whether everyday users actually adopt these tools — or whether, as one Google executive admitted on stage, agentic AI is still in its “early days.”
What was your favorite Google I/O 2026 announcement? Drop a comment below — and bookmark InsiderXP for daily tech, AI, and gadget news.
FAQ
- When was Google I/O 2026 held?
May 19–20, 2026, at Shoreline Amphitheatre, Mountain View, California. - What is Gemini Spark?
A new “24/7” personal AI agent in the Gemini app that runs on Google Cloud and can complete tasks even when your device is off. - When will Samsung Intelligent Eyewear release?
Fall 2026, with designs by Warby Parker and Gentle Monster. Pricing has not yet been announced. - How much does Google AI Ultra cost?
$100 per month, aimed at developers, creators, and power users.
