Google I/O 2026: From AI agents to smart glasses, here are the biggest announcements | Technology News

Gemini 3.5 Flash with faster AI agents and coding tools

Google has introduced Gemini 3.5, its latest family of AI models focused on combining “frontier intelligence” with advanced agentic capabilities. The first release, Gemini 3.5 Flash, is designed for complex multi-step workflows, coding, automation, and real-world AI tasks while delivering significantly faster performance than rival frontier models. Google says the model excels in coding, multimodal reasoning, and long-horizon task execution, with applications ranging from app development and financial document preparation to enterprise automation.

Gemini 3.5 Flash combines frontier intelligence with action capabilities, delivering outputs up to four times faster than rival AI models.

Gemini 3.5 Flash also powers new AI experiences across the Gemini app, Google Search, and the upcoming Gemini Spark personal AI agent. The company highlighted stronger safety protections, enhanced cyber safeguards, and improved AI transparency. Gemini 3.5 Flash is now available globally through the Gemini app, AI Mode in Search, Google AI Studio, Android Studio, and enterprise platforms, while Gemini 3.5 Pro is expected next month.

Gemini app gets proactive AI agents, cinematic video tools and redesigned interface

Google has announced a major upgrade to the Gemini app at I/O 2026, introducing more proactive AI-powered features, a redesigned interface, and new agentic capabilities. The update includes Gemini 3.5 Flash, Google’s latest fast-performing AI model, alongside a refreshed “Neural Expressive” design language with fluid animations, richer visuals, and integrated conversational voice interactions. Google also unveiled Gemini Omni, a new multimodal AI tool capable of generating and editing cinematic-quality videos using text, image, and video prompts. Another key addition is Daily Brief, an AI agent that creates personalised morning summaries using Gmail, Calendar, and other connected apps. Google also introduced Gemini Spark, a cloud-based 24/7 AI agent that can automate recurring tasks, organise workflows, monitor emails, and interact with connected services like Canva and OpenTable. Many of the new features will initially roll out to Google AI Plus, Pro, and Ultra subscribers, while a macOS Gemini desktop app with advanced voice features is also launching later this year.

Google introduces AI-powered Search with agents, smarter browsing and custom mini apps

Google has announced a major overhaul of Search at I/O 2026, introducing new AI-powered features designed to make Search more conversational, proactive, and agentic. The company is upgrading AI Mode with Gemini 3.5 Flash as the default model globally, alongside what it calls the biggest redesign of the Search box in over 25 years. The new AI-powered Search box supports text, images, videos, files, and Chrome tabs while offering contextual suggestions and conversational follow-ups. Google is also introducing “Search agents,” which can continuously monitor the web, news, shopping, finance, and sports updates for users and send personalised alerts or recommendations.

(Image: Google)

Search is additionally gaining agentic booking tools for services, local experiences, and shopping tasks. Google also previewed AI-generated mini apps and dashboards powered by Google Antigravity, enabling Search to build custom trackers, visual simulations, and interactive tools on demand. Expanded Personal Intelligence support will also allow users to securely connect Gmail, Photos, and Calendar for more context-aware Search experiences across nearly 200 countries and 98 languages.

Gemini Spark: Google’s 24/7 AI agent for everyday tasks

Google has unveiled Gemini Spark, a new always-on AI agent designed to move Gemini beyond answering questions and into actively completing tasks on a user’s behalf. Powered by Gemini 3.5 and Google’s Antigravity framework, Spark integrates deeply with Workspace apps like Gmail, Docs, and Slides, allowing it to automate recurring workflows, summarise meeting notes, monitor inboxes, and generate reports or emails in the background. Google says Spark can continue working even when a device is locked or offline. The company is also expanding integrations with services like Canva, OpenTable, and Instacart. Spark will begin rolling out to trusted testers and Google AI Ultra subscribers in the US.

Gemini Omni Flash for AI video generation and editing

Google has introduced Gemini Omni, a new multimodal AI model family designed to generate and edit content from text, images, audio, and video inputs, starting with video creation. The first model, Gemini Omni Flash, is rolling out to the Gemini app, Google Flow, YouTube Shorts, and YouTube Create. Google says Omni combines Gemini’s reasoning abilities with generative media tools, allowing users to edit videos conversationally while maintaining scene consistency, physics, and character continuity across multiple prompts.

The model can generate videos grounded in real-world knowledge, create explainers, apply styles and effects, and combine multiple input formats into cohesive outputs. Google is also introducing AI-generated digital avatars based on users’ voices and likenesses, while embedding SynthID watermarks into all AI-generated videos for transparency and verification.

Google expands Antigravity with faster Gemini 3.5 Flash and AI coding tools

Google has announced major developer-focused upgrades centred around Gemini 3.5 Flash, its new high-speed AI model designed for agentic workflows and coding tasks. The company says the model delivers frontier-level performance while operating four times faster than rival frontier AI systems. Alongside the model, Google introduced Antigravity 2.0, a standalone desktop app for orchestrating multiple AI agents, as well as new command-line and SDK tools for developers. Google is also launching Managed Agents in the Gemini API, allowing developers to deploy AI agents capable of reasoning, using tools, and executing code in isolated environments with a single API call.

Antigravity 2.0 updates aim to give developers more powerful tools and greater control over AI agents.

The company further expanded Google AI Studio with native Android app generation, Workspace integrations, Play Console publishing support, and a new mobile app for rapid prototyping on the go. Google also announced a new $100-per-month AI Ultra subscription tier with higher usage limits for Antigravity tools. To encourage AI development, the company unveiled the Build with Gemini XPRIZE Hackathon, featuring a $2 million prize pool for developers building AI-powered applications using Gemini models and tools.

Google introduces Universal Cart for AI-powered shopping

Google has unveiled Universal Cart at Google I/O 2026, an AI-powered shopping hub designed to make online purchases smarter and more proactive. Powered by Gemini and Google’s Shopping Graph of over 60 billion product listings, Universal Cart works across Search, Gemini, YouTube, and Gmail, automatically tracking deals, price drops, stock alerts, and compatibility issues for products added to the cart. Google says the system can even recommend alternatives and identify hidden savings using Google Wallet data like loyalty rewards and payment perks. The company also expanded its Universal Commerce Protocol (UCP) and introduced the Agent Payments Protocol (AP2), enabling secure AI-assisted purchases with user-defined spending limits and verification safeguards. Universal Cart will begin rolling out in the U.S. this summer.

Google Workspace gets Gemini voice tools, AI Inbox and new image editor Google Pics

Google has announced a major set of AI-powered upgrades for Workspace at I/O 2026, introducing new voice-driven productivity tools, an AI image editor called Google Pics, smarter inbox management, and deeper Gemini integration. Gmail, Docs, and Keep are gaining conversational “Live” features that let users search emails, brainstorm documents, and organise notes using natural voice commands. Google also unveiled Google Pics, a new AI image generation and editing app powered by Nano Banana, offering object editing, text translation inside images, collaborative canvases, and direct integration with Slides and Drive.

Gmail’s AI Inbox is expanding with contextual draft replies, surfaced file links, and improved task management tools. Google additionally confirmed that Gemini Spark, its new 24/7 AI agent, will integrate with Workspace apps to automate tasks and assist with emails, documents, and scheduling under user control.

Daily Brief: Gemini’s new AI morning assistant

Google has introduced Daily Brief, a new Gemini-powered AI agent designed to deliver a personalised morning summary of everything users need to know for the day. Once enabled, Daily Brief works across connected apps like Gmail and Calendar to gather urgent emails, upcoming events, reminders, and relevant follow-up information into a single easy-to-read briefing. Unlike a standard summary tool, it actively prioritises tasks based on user goals and even suggests immediate next steps. Users can also refine recommendations over time using simple feedback controls. Daily Brief is rolling out first to Google AI Plus, Pro, and Ultra subscribers in the US.

Google teases Gemini-powered smart glasses launching later this year

At Google I/O 2026, Google unveiled its upcoming intelligent eyewear platform built on Android XR, introducing AI-powered smart glasses developed in partnership with Samsung, Gentle Monster, and Warby Parker. The first products will be audio-focused smart glasses arriving later this fall, with future models also featuring built-in displays. Powered by Gemini AI, the glasses allow users to get directions, send texts, manage calls, translate speech in real time, snap and edit photos, and interact with apps using voice commands without needing to take out their phone.

(Image: Google)

Google says the glasses can understand what users are looking at, provide contextual information about surroundings, and assist with tasks such as food ordering, navigation, and ride booking. Features include hands-free messaging, live translations, camera-based AI queries, and integrations with apps like Uber, DoorDash, and Mondly. The glasses will support both Android and iOS devices as Google expands its vision for always-available AI-powered wearable computing.

Google unveils Docs Live AI voice-writing tool at I/O 2026

At Google I/O 2026, Google introduced Docs Live, a new AI-powered feature for Google Docs that can transform spoken thoughts into structured written content. Designed as a conversational writing assistant, the tool lets users freely dictate ideas while AI organizes them into outlines, drafts, tables, and formatted documents. With user permission, Docs Live can also pull context from connected Google services like Gmail, Drive, and Chat, as well as information from the web, to refine generated text. Google positions the feature as a faster, more natural way to brainstorm and draft content using voice rather than prompts. Docs Live will launch this summer for Google AI Pro and Ultra subscribers.

Google redesigns Gemini with ‘Neural Expressive’ AI interface

Google has unveiled Neural Expressive, a major redesign of the Gemini experience focused on making AI interactions more dynamic and conversational. The new interface introduces fluid animations, vibrant visuals, updated typography, and haptic feedback across web, Android, and iOS. Google has also integrated Gemini Live directly into the main Gemini experience, allowing users to switch seamlessly between typing and natural voice conversations. The updated microphone system is designed to better handle free-flowing speech without interruptions, while future updates will add support for regional dialects. Gemini responses are also becoming more interactive, with AI-generated visuals, timelines, narrated videos, and dynamic graphics replacing traditional text-heavy outputs.

Stitch with real-time AI design collaboration

Google has announced major upgrades to Stitch at I/O 2026, turning the AI design tool into a real-time collaborative platform powered by the Stitch Agent. Users can now create and refine designs using text prompts, voice commands, existing codebases, or design files, making the workflow more natural and interactive. Stitch can instantly generate shareable links through Google AI Studio, while completed projects can be exported to Google Antigravity for backend integration or published directly to the web using Netlify. Google says the new features are aimed at simplifying rapid prototyping and live collaborative design.

Managed Agents in Gemini API simplify AI agent development

Google has introduced Managed Agents in the Gemini API, allowing developers to launch AI agents with a single API call using the Antigravity agent powered by Gemini 3.5 Flash. The system runs in isolated Linux cloud environments where agents can reason, use tools, browse the web and execute code. Developers can also create custom agents using markdown-based AGENTS.md and SKILL.md files instead of complex orchestration code. Sessions remain persistent across interactions, preserving files and context. Managed Agents will be available through the Gemini API, Google AI Studio and Gemini Enterprise Agent Platform, helping developers build scalable production-grade AI agents faster and with less infrastructure overhead.