Live from Shoreline: An AI-First Google I/O Keynote
Streaming from the Shoreline Amphitheatre in Mountain View, Google I/O 2026 opened with a clear message: Gemini now underpins almost everything Google builds. The main keynote, followed by a developer-focused session, mixed live demos, hardware teasers, and deep dives into new AI-driven experiences across phones, laptops, and wearables. Attendees and online viewers watched as executives framed this year’s event around agent-first computing, promising Gemini-powered help in Search, Workspace, Chrome, and new creative tools. Google highlighted recent Android updates and the broader push toward multimodal, always-available assistance that can see, hear, and act on users’ behalf. With on-site reporting and companion watch parties running on community Discord servers, the keynote set the tone for two days of codelabs, sessions, and hands-on time with the latest Gemini models, Android XR hardware, and upgraded developer tooling.
Gemini 3.5 Flash, Omni, and the AI-First Product Stack
On the model front, Google used the keynote to spotlight Gemini 3.5 Flash as its next big upgrade, finally giving stage time to the speed-focused model after it briefly appeared to early users under the “Gemini 3 Fast” label. Alongside Flash, Gemini Omni emerged as the centerpiece of Google’s multimodal story, designed to unify text, image, and video generation within a single architecture and now surfacing in products like Google Flow and the Gemini app. The company reiterated that Gemini will power everything from an upgraded Google Search bar with richer generative answers to new creative tools for images and video. While Gemini 3.5 Pro remained more of a teaser, Google positioned its growing family of models as the backbone for agentic experiences, scientific research tools, and AI assistants that can reliably understand complex tasks and operate across many surfaces.
Android XR Glasses and Intelligent Eyewear Take Center Stage
Hardware finally shared the spotlight when Google and Samsung showed off the first consumer-ready Android XR glasses, developed in partnership with eyewear brands Warby Parker and Gentle Monster. Described as intelligent eyewear, these Android XR glasses are slated to arrive later this year in multiple styles and finishes. Google also previewed its first audio-based smart glasses, planned for launch in the fall, designed to keep Gemini assistance accessible without pulling out a phone. In live demos, the AI-powered glasses recognized what the wearer was seeing, controlled music playback, and even handled a coffee order via DoorDash, only asking for final confirmation before purchase. Together, these announcements framed Android XR glasses as a practical extension of Gemini, bringing heads-up navigation, contextual understanding, and hands-free actions into everyday routines while keeping the classic look of regular frames.
A Redesigned Gemini Experience and New Developer Tools
Beyond models and hardware, Google unveiled a fully refreshed Gemini app, described as a complete redesign built in a visual language called Neural Expressive. The new interface introduces updated typography, colors, and Gemini Live, which lets users fluidly switch from typing to talking while supporting more languages and regional experiences. On desktop, the Gemini app is gaining Spark, an agent mode that can work with local folders, connectors, and custom skills. Developers also get a boost from Google AI Studio’s upcoming mobile companion, enabling coding directly from phones, plus new Modern Web Guidance and Chrome DevTools aimed at helping coding agents build and debug modern web apps. Additional tools like WebMCP allow developers to turn ordinary web pages into agent toolkits, pushing Gemini further into the workflow of building, testing, and deploying AI-powered applications.
Gemini for Science and the Future of Agentic AI
To close the keynote, Google shifted from consumer features to long-term impact, introducing Gemini for Science as a dedicated initiative to accelerate research. The suite of tools is designed to help scientists track new papers, summarize updates, and stay current across fast-moving fields, with ambitions that span medical research, drug discovery, climate modeling, and weather prediction. Executives described how Gemini can help build digital twins of the Earth to simulate real-world scenarios and inform better decisions. This scientific push sits alongside expansions in creative tools like Google Flow for filmmakers and Pics for image editing, both powered by Gemini’s latest models. Framed as just the “foothills” of a much bigger AI transformation, the keynote made clear that Google sees agentic AI not only reshaping apps and Android XR glasses, but also driving breakthroughs in how knowledge is created and shared.
