Google I/O 2026 Recap: Gemini 3.5, Gemini Omni, Spark and Everything That Shipped

What Actually Shipped

Google I/O 2026 ran 19–20 May at the Shoreline Amphitheatre. This article was originally published as a preview on 17 May. What follows is what actually happened.

The headline: this was the most agent-heavy I/O yet. Nearly every major announcement involved an AI that runs in the background, asks before acting, and surfaces results when ready.

Gemini 3.5 Flash — Available Now

Gemini 3.5 Flash is the first model in the 3.5 family and it is generally available via Google AI Studio, the Gemini API, Android Studio, and Antigravity. Google positions it for fast agentic work, coding, and long-horizon tasks.

Gemini 3.5 Flash is also now the default model in Google AI Mode (Search), which Google describes as the biggest Search upgrade in 25 years — users can now search across text, images, files, videos, and Chrome tabs.

Gemini 3.5 Pro was teased but not released. Google says it is being used internally and is expected to roll out in June 2026.

Gemini Omni — Video Generation Rebranded

Gemini Omni is the new multimodal generative model that replaces what was previously called Veo. It starts with video but is pitched as “any output from any input” over time. It combines Gemini intelligence with Google’s generative media models and includes SynthID watermarking.

Gemini Omni Flash is rolling out to consumers and creators through the Gemini app and Google Flow for Google AI Plus, Pro, and Ultra subscribers. It also appears in YouTube Shorts Remix and YouTube Create for users 18+, with conversational video editing, background changes, templates, and improved character consistency.

Gemini Spark and the Agent Wave

Gemini Spark is a 24/7 personal AI agent powered by Gemini 3.5 and built on Antigravity. It works in the background, asks before taking major actions, and will later support texting or emailing the agent, custom sub-agents, and authorised payments.

Daily Brief is a Gemini app agent that works overnight to analyse connected Gmail, Calendar, and tasks, then delivers a prioritised digest with suggested next steps. Rolling out first to Google AI subscribers in the US.

Search information agents are background agents that monitor web, news, social, finance, shopping, and sports updates and send synthesised notifications. Due this summer for Google AI Pro and Ultra subscribers.

Antigravity 2.0 — Agent Development Platform

Antigravity 2.0 expanded from a coding tool into a full agent-first development platform with a standalone desktop app, CLI, SDK, subagents, hooks, async task management, and multi-agent orchestration. Google says AI Studio can now build native Android apps from prompts and publish to a Google Play internal test track.

Everything Else Worth Noting

• Neural Expressive: A new Gemini interface style with fluid animations, richer layouts, haptics, embedded visuals, timelines, and a more inline Gemini Live experience.
• Universal Cart: An AI shopping cart that gathers items from Search, Gemini, YouTube, and Gmail, monitors price drops, flags incompatibilities, and uses Google Wallet perks. Rolling out this summer.

How to Evaluate These Announcements

A “real” announcement answers five questions:

1. Availability: When can I actually use this?
2. Location: Where is it available?
3. Price: How much does it cost?
4. Users: Who is this for?
5. Workflow change: How does this change what I do on Monday morning?

By that standard, Gemini 3.5 Flash and Gemini Omni Flash are real products. Gemini Spark, Daily Brief, Search agents, and Universal Cart are rolling out — watch for actual availability in your region. Gemini 3.5 Pro is still a preview.

The Mastering AI Tools course is being updated with the new Google AI features. Check the Resources page for quick-reference guides.

Frequently Asked Questions

What was announced at Google I/O 2026?

Key announcements include Gemini 3.5 Flash (generally available), Gemini Omni (multimodal video generation replacing Veo), Gemini Spark (personal AI agent), Antigravity 2.0 (agent development platform), Daily Brief, Search information agents, Universal Cart, and a Gemini redesign with Neural Expressive.

Is Gemini 3.5 Pro available?

Gemini 3.5 Pro was teased but not released at I/O. Google says it is being used internally and is expected to roll out in June 2026.

What is Gemini Omni?

Gemini Omni is Google’s new multimodal generative model that replaces Veo. It starts with video but is designed for any output from any input. Gemini Omni Flash is available to Google AI Plus, Pro, and Ultra subscribers through the Gemini app and Google Flow.