After the excitement around the opening day of Google I/O, Day Two was focused on the practical. Gemini 3.5 Flash, Gemini Omni, AI Search, Universal Cart, Gemini Spark, Antigravity, Workspace updates, Android XR and SynthID all show Google pushing AI deeper into work, shopping, search, creativity and wearable hardware.
Day 1: Keynote, Day 2: Product Map
If Google I/O Day One gave Android XR glasses its moment, Day Two showed how wide Google wants the Gemini push to go.
The company’s official roundup lists 100 announcements across models, Search, shopping, the Gemini app, developer tools, Workspace, creative products, science, YouTube and Android XR .
It’s a lot, even by Google standards!
The direction in Google I/O is clear enough. Google is pushing Gemini out of the chatbot box and into products people already use daily: Search, Gmail, Docs, Android Studio, YouTube, shopping, creative tools, scientific work and glasses.
Highlights:
- Gemini 3.5 Flash gives Google a faster model for agentic work
- Gemini Omni pushes video and multimodal creation forward
- Search gets information agents, generated interfaces and custom trackers
- Universal Cart moves AI closer to shopping decisions and checkout
- Gemini Spark brings a personal agent into the Gemini app
- Android XR keeps glasses and spatial hardware inside the Gemini roadmap
Gemini Splits Between Action and Creation
Google’s model updates can broadly be split between practical, action-focused, and creative.
Gemini 3.5 Flash is the probably the main action-focused update. Google says it is the first model in its latest Gemini 3.5 series combining frontier intelligence with action, and it is generally available through Google Antigravity, the Gemini API in Google AI Studio and Android Studio.
Google has said:
“Gemini 3.5 Flash is ideal for tackling long-horizon agentic tasks. What used to take a developer days or an auditor weeks, 3.5 Flash can now help complete in a fraction of the time, often at less than half the cost of other frontier models.”
It looks like google wants Gemini to take on a multitude of new tasks: planning, building, iterating and completing longer workflows across tools, rather than sitting inside one prompt and one answer.
Gemini Omni is the more creative update:
“Gemini Omni is our new model that can create anything from any input — starting with video. It combines Gemini’s intelligence with the best of our generative media models for a new level of world understanding, multimodality and editing.”
Google is folding creation, editing, remixing, avatars, video and generated tools into the same AI product cycle. For creators and marketing teams, the appeal is obvious., but businesses, will have questions around rights, brand safety, workflow control and how quickly synthetic media gets reviewed before it reaches customers.
Google is Trying to Redefine What “Googling” Something Means
Search picked up some of the biggest product changes on Day Two.
Google says AI Mode has passed more than 1 billion monthly users, and Gemini 3.5 Flash is becoming the new default model globally for the experience. Google also described a reworked Search box that can handle text, images, files, videos and Chrome tabs.
There’s an intereting update on the agent layer too.
“Your agent will intelligently look across everything on the web, like blogs, news sites and social posts, plus our freshest data, such as real-time info on finance, shopping and sports, to monitor for changes related to your specific question.”
Google says information agents will roll out this summer, first to Google AI Pro and Ultra subscribers.
Search will also generate custom interfaces. Google says users will be able to get layouts with interactive visuals, tables, graphs or simulations, plus longer-running custom experiences such as dashboards and trackers.
Shopping then gets pulled into the same lane through Universal Cart. Google says users will be able to add products while browsing Search, chatting with Gemini, watching YouTube or reading Gmail, with the cart tracking deals, price drops, stock changes, product compatibility and payment perks.
Universal Cart is rolling out across Search and the Gemini app this summer, with YouTube and Gmail to follow.
Spark And Workspace Move Agents Into Daily Work
The Gemini app also moved closer to a personal agent.
Google describes Gemini Spark as a personal AI agent that can work in the background, act under the user’s direction and check before major actions. Spark runs on Gemini 3.5 and is built on Antigravity. Google says it is rolling out first to trusted testers, with a beta planned for Google AI Ultra subscribers in the U.S. next week.
Spark takes the Gemini app into more sensitive territory: background work, delegated tasks, future payments, custom sub-agents and direct access through text or email later this summer. Google is moving carefully, but the direction is clear.
Workspace also picked up practical updates. AI Inbox is expanding beyond Ultra subscribers to Google AI Plus and Pro subscribers in the U.S., with personalized draft replies, surfaced links to Docs, Sheets and Slides, and cleaner task management. Gmail Live will let Google AI Pro and Ultra subscribers talk to their inbox from this summer.
Docs Live will let users create and edit documents by voice, with permission-based access to Gmail, Drive, Chat and the web. Keep is also getting a feature that turns spoken thoughts into organized notes and lists.
For workplace tech teams, these updates are easy to picture. Drafting, sorting, finding, organizing and acting across existing tools is easier than asking users to learn a totally new work surface.
Antigravity Gives Builders More Agent Firepower
Google also pushed Antigravity harder.
Antigravity 2.0 is now a standalone desktop application for working with multiple agents in parallel. Google also announced Antigravity CLI, Antigravity SDK and Managed Agents in the Gemini API.
AI Studio is getting more practical too. Google says developers can build native Android apps in the build tab, connect Google Play Console, preview apps through an Android Emulator in the browser, deploy early apps to Google Cloud and export work into Antigravity.
That connects neatly to the XR and wearable side. Android XR needs more than nice hardware on stage. It needs apps, tools, testing routes and developers who can build without fighting the platform at every step.
Meta has just opened Ray-Ban Display to developers , while Google is trying to make Android XR part of a wider Gemini and Android toolchain. The platform fight around smart glasses is already moving into apps, input, agents and developer access.
Android XR Stays In The Mix
Android XR did not dominate Day Two in the same way glasses did on Day One, but it stayed in Google’s official roundup.
Google repeated that intelligent eyewear is the next major Android XR milestone, with audio glasses offering spoken help and display glasses showing information in the user’s line of sight .
The first audio glasses, made with Gentle Monster, Warby Parker and Samsung, are due this fall and will work with Android and iOS devices.
Android XR now has an eyewear path, a partner list, an autumn window and a clearer role inside Google’s Gemini push.
Workplace buyers will still need answers on management, app approval, data capture, privacy controls and whether the hardware can survive normal working condition]. Glasses with cameras, microphones and AI assistance will face tougher checks in offices, schools, public spaces and regulated environments than they will in a keynote demo.
What This Means For XR
- Android XR is now part of Google’s wider Gemini push
- Glasses give AI a route beyond phones, laptops and tablets
- Search agents and Spark show how Google wants AI to act across longer tasks
- Antigravity and AI Studio matter because device platforms need developers
- SynthID points to the trust problem facing AI video, spatial media and synthetic content
- Workplace buyers will need clear answers on management, privacy and approval workflows
Day Two Made Google’s AI Ambition Harder To Ignore
Day One put Android XR glasses back on the calendar, but Day Two showed how much else Google wants Gemini to touch.
Search gets agents. Shopping gets Universal Cart. Workspace gets more automation. Developers get Antigravity, Managed Agents and stronger AI Studio tooling. Creators get Omni, Flow, Pics and more video editing firepower. Android XR keeps glasses and spatial hardware inside the product roadmap.
Google has no shortage of AI announcements now. The next test is whether these agents, tools and devices work well enough for people to trust them with everyday tasks, workplace documents, shopping decisions, creative work and AI wearables that can see and hear what’s around them.
Gemini is being pushed into the operating layer of digital life. In terms of XR, it’s encouraging that glasses and spatial hardware are now part of that plan. Let’s wait and see what Google I/O Day 3 brings!
