Google I/O 2026: What You Can Use Today vs. What's Still Coming
May 19, 2026 | AI Strategy
Google's I/O keynote this morning packed in more announcements than any single session could absorb. Gemini 3.5 Flash, a new general-purpose agent called Gemini Spark, Managed Agents in the developer API, Android XR smart glasses, and a new shopping layer called Universal Cart. The challenge with I/O is that "announced" and "available" are two very different things. Here is the breakdown.
What Ships Today
Gemini 3.5 Flash is the one that matters most for practitioners right now. Google says it outperforms Gemini 3.1 Pro across nearly all benchmarks while running four times faster and at roughly half to one-third the price. It is rolling out today in the Gemini app, Search, and the Gemini API. For anyone building on the Gemini API, this is a direct upgrade path: faster output tokens per second, better agentic task performance, lower cost. That combination is what makes it usable in production rather than just impressive in demos.
Gemini Omni is also live today. It is a new model series that accepts image, audio, video, and text input and generates video output grounded in real-world knowledge. Think of it as a creation model layered on top of Gemini's reasoning. Available in the Gemini API starting today.
Google AI Studio and Antigravity updates are shipping now for developers. The Antigravity platform is Google's agent-first development environment, and the expanded toolset lets developers build, manage, and deploy agents that integrate across key developer surfaces. If you are building production agents, this is worth a serious look today.
What Is Coming Soon
Gemini Spark is the announcement that grabbed the most attention, and it is not fully available yet. Spark is Google's new general-purpose agent in the Gemini app. It can reason across connected apps, take action on your behalf, and navigate multi-step tasks autonomously. It launches in beta next week for trusted testers and Google AI Ultra subscribers. Broader availability has no confirmed date.
Managed Agents in the Gemini API are coming soon for developers. A single API call will spin up an agent that reasons, uses tools, and executes code in an isolated Linux environment. The infrastructure layer that makes agentic systems faster to build. No ship date beyond "coming soon."
Gemini 3.5 Pro is expected next month. Google confirmed it is in testing.
Generative UI in Search is shipping this summer, free for everyone. Google is using Antigravity and Gemini 3.5 Flash to build the response format on the fly based on your query. Ask about astrophysics and you get an interactive simulation. Plan a wedding and Search builds you a custom dashboard you can return to and update. These are not static results pages. They are generated mini-apps rendered directly in Search. Available this summer.
Android XR intelligent eyewear is the hardware story. Two product lines: Audio Glasses and Display Glasses, both powered by Android XR and Gemini. Samsung built them in partnership with Gentle Monster and Warby Parker for the design. Audio Glasses launch first, this fall. Pricing and specific regional availability have not been announced.
Universal Cart is Google's AI-powered shopping layer. Intelligent cart experience that works across retailers. No availability date.
What This Means for Business Builders
The Gemini 3.5 Flash release is the most actionable item from today. If you are evaluating AI models for production use and cost is a constraint, the benchmark performance combined with the price reduction makes it a strong candidate to test this week.
The Gemini Spark announcement is the signal worth watching closely. A general-purpose agent that acts on your behalf across connected systems is the same infrastructure layer that businesses are going to expect their vendors to build on. The pattern I have been building into The Latent Space is not unique to my use case. It is the direction the entire stack is moving. When Google ships a consumer version of this concept, it accelerates adoption expectations for everyone.
The generative UI announcement is the one that should make web builders pay attention. When Search starts rendering custom interactive dashboards instead of links, the traffic pattern changes. Users who would have clicked through to your site to track something will stay in Search and do it there. That is not a future problem. It is a this-summer problem for anyone whose business depends on return visits to a content or tool page.
The Android XR glasses are a fall story, not a today story. Keep an eye on how the Gemini integration works in real-world testing as reviews come in. The audio-first approach suggests they learned from earlier AR hardware: get the use case working before getting the display working.
One thing I have learned building production AI systems: the announcements that move fast are always the ones with a developer API attached to them at launch. Gemini 3.5 Flash and Omni have that today. Gemini Spark does not yet. That gap is usually where the 6-to-12-month wait lives.
Sources: Google I/O 2026 announcements, Google Blog. Gemini 3.5 Flash and Spark, CNBC. Android XR eyewear release, 9to5Google. Generative UI in Search, Interesting Engineering.
Written by Travis Raveling, Founder PAID LLC, co-authored and edited by AI.
About PAID LLC: PAID LLC helps businesses understand, implement, and get ROI from AI tools and emerging technology. Learn more at paiddev.com/about.