Catch Up on 12 Major I/O 2026 Moments

Key Takeaways

→ Gemini Omni Flash generates and edits video from combined image, audio, video and text input

→ Gemini 3.5 Flash targets agents and coding; Gemini 3.5 Pro is promised for the following month

→ Information agents run in the background 24/7 and roll out first to Google AI Pro and Ultra subscribers

→ Search gains Antigravity-powered generative UI and custom mini apps for long-running tasks

→ SynthID has watermarked over 100 billion images and videos plus 60,000 years of audio

→ Audio smart glasses launch in fall 2026; Gemini for Science links Antigravity to 30+ life science databases

Google used its annual I/O developer conference in May 2026 to launch two new model families and a sweeping set of agentic features across Search, Gemini and Android. In a May 28 recap, the company collected 12 of the most significant moments from the event.

Gemini Omni and Gemini 3.5 Flash

Gemini Omni is a model that creates anything from any input, starting with video. Users can mix images, audio, video and text, then edit the result through conversation. The first family member, Gemini Omni Flash, rolled out globally to Google AI Plus, Pro and Ultra subscribers through the Gemini app and Google Flow, and free to users on YouTube Shorts and the YouTube Create app.

Gemini 3.5 Flash opens a new frontier family built around action. It is generally available through Google Antigravity, the Gemini API, Android Studio and Gemini Enterprise, and is rolling out in the Gemini app. Google said Gemini 3.5 Pro was already in internal use and would follow the next month.

Agents Arrive in Search

Google introduced information agents that reason across blogs, news and fresh data around the clock, plus Antigravity-powered experiences that build dynamic layouts, dashboards and mini apps directly inside Search. Users create an information agent by adding keep me updated to a query.

Daily Brief, Universal Cart and Spark

Daily Brief compiles email, calendar and follow-ups into a morning briefing in the Gemini app
Universal Cart tracks prices, history and stock across merchants and Google services
Gemini Spark is a cloud-based agent that acts across Gmail, Docs and Slides and asks before high-stakes actions

A New Look and New Devices

Neural Expressive redesigns the Gemini experience with fluid animation, new typography and responses that Gemini designs in real time. Google also detailed a macOS app with local-file automation and two types of intelligent eyewear, with audio glasses arriving in fall 2026.

Provenance Gets Stronger

SynthID verification is expanding to Search and Chrome in the coming weeks, and Google is launching an AI content detection API on Google Cloud. Content Credentials are extending to video on Pixel 8, 9 and 10 phones.

Gemini for Science

Gemini for Science bundles Deep Think and Deep Research with Science Skills that connect Google Antigravity to more than 30 major life science databases and tools.

What This Means

I/O 2026 was Google’s clearest statement yet that its strategy is agents everywhere: in Search, in the Gemini app, on the desktop and in hardware. The company is betting that distribution across its existing products, not model quality alone, wins the assistant era.

Frequently Asked Questions

What was announced at Google I/O 2026?

Google announced the Gemini Omni video model, the Gemini 3.5 Flash model, information agents and generative UI in Search, the Daily Brief and Gemini Spark agents, the Neural Expressive redesign, a macOS Gemini app, intelligent eyewear, a SynthID expansion into Search and Chrome, and Gemini for Science.

What is Gemini Omni?

Gemini Omni is a Google model that creates content from any input, starting with video. Users can combine images, audio, video and text and then edit the result by conversation. The first release, Gemini Omni Flash, rolled out to AI Plus, Pro and Ultra subscribers and free on YouTube Shorts.

Which I/O features are available first to paying subscribers?

Information agents, Antigravity-powered experiences in Search, Daily Brief and Gemini Spark all arrive first for Google AI Pro and Ultra subscribers, with broader rollouts promised later in 2026.

How many assets has SynthID watermarked?

Google said SynthID has embedded imperceptible watermarks in more than 100 billion images and videos and 60,000 years of audio assets. Verification is expanding to Search and Chrome, and OpenAI, Kakao and ElevenLabs are adopting the technology.