100 Things Google Announced at I/O 2026

Introduction

Google I/O 2026 was not a subtle event. Furthermore, Google made that clear by publishing a list of exactly 100 things it announced across the two-day conference held on May 19 and 20 at the Shoreline Amphitheatre in Mountain View, California. Consequently, this guide works through every category of announcement from new frontier AI models and 24/7 personal agents to cross-platform shopping carts, smart glasses, scientific research tools, and custom silicon so that developers, marketers, professionals, and everyday users can understand the full scope of what changed at Google I/O 2026 and what it means for the road ahead. Google CEO Sundar Pichai described the moment from the stage as the beginning of the "agentic Gemini era" a period where AI no longer waits to be asked but actively works on behalf of users around the clock. Every announcement at the conference reflects some dimension of that transition.

The Complete Numbered List: All 100 Things Google Announced at I/O 2026

Gemini Models (1-6)

1. Gemini 3.5 Flash launched as the first model in Google's latest series combining frontier intelligence with genuine agentic capability across coding, multimodal, and long-horizon task benchmarks.

Gemini Omni Capabilities (7-16)

7. Gemini Omni Flash is the first model in the Omni series - available on May 19 for all Google AI Plus, Pro, and Ultra subscribers globally through the Gemini app and Google Flow.

8. Gemini Omni Flash supports conversational video editing - users apply cinematic zooms, swap backgrounds, and change characters through natural spoken prompts without a traditional timeline.

9. Gemini Omni can simulate real-world physics including kinetic energy, gravity, and material behavior in generated video content.

10. Gemini Omni translates complex scientific concepts into visually comprehensible video output, making dense academic content accessible.

11. Gemini Omni supports personal AI avatar creation - users can generate videos in their own voice and likeness using their personal reference data.

12. Every Gemini Omni output carries an imperceptible SynthID watermark verifiable through the Gemini app, Gemini in Chrome, and Google Search.

13. Gemini Omni turns any reference into output - image, text, video, or audio references produce a single cohesive generated output.

14. Gemini Omni is available at no cost for users aged 18 and older through YouTube Shorts Remix and the YouTube Create App from May 19.

15. YouTube Shorts Remix with Gemini Omni lets users step directly into any eligible Short by prompting a change - adding themselves or any visual reference to produce a fresh version.

16. Gemini Omni developer and enterprise API access arrives in the weeks following the keynote, enabling production integrations across third-party platforms.

Custom Silicon: TPU 8 (17-20)

17. TPU 8t - Google's eighth-generation training chip - delivers nearly three times the raw compute of the previous generation with up to two times better performance per watt.

18. TPU 8i - the companion inference chip - enables globally distributed inference workloads, running AI model responses closer to users around the world.

19. Google can now distribute AI training across more than one million TPUs globally using JAX and Pathways infrastructure, creating the world's largest training cluster.

20. Training large, capable models now takes weeks rather than months with the new multi-site training architecture enabled by TPU 8t and the Pathways system.

Gemini App: Redesign and Features (21-26)

21. Neural Expressive is the Gemini app's new design language - introducing fluid animations, vibrant responsive colors, haptic feedback, and revised typography throughout.

22. The Gemini app moves to a compute-based usage model - interactions consume resources proportional to their actual computational demands rather than counting against a fixed daily prompt cap.

23. Usage limits now refresh every five hours and count against a weekly compute budget, distributing capacity more equitably across different task types and user patterns.

24. Regional dialect support for voice output allows users to select localized speech variations within their primary language for more contextually natural audio responses.

25. Personal Intelligence expands to nearly 200 countries across 98 languages - connecting Gmail and Google Photos to AI Mode for personalized, context-aware answers at a global scale.

26. The Gemini app becomes more agentic with Gemini Omni, Gemini Spark, and Daily Brief all integrated directly into the redesigned Neural Expressive interface.

Gemini Spark: 24/7 Personal Agent (27-33)

27. Gemini Spark is a 24/7 personal AI agent that runs entirely on Google Cloud, working on the user's behalf even when their device is locked, closed, or offline.

28. Gemini Spark connects to Gmail, Google Docs, Sheets, and Workspace at launch, with third-party app support via MCP arriving over the summer.

29. Gemini Spark parses credit card statements to identify overlooked recurring subscriptions and flag unnecessary charges automatically.

30. Gemini Spark monitors email inboxes for deadlines, school notices, and work communications - flagging time-sensitive items and surfacing them proactively.

31. Gemini Spark transcribes meeting notes into Google Docs and emails the formatted summary to all attendees without manual input.

32. Gemini Spark drafts project emails from brief voice notes - converting a spoken summary into a complete, ready-to-send message.

33. The Agents Payment Protocol in Spark enables fully automated purchasing within spending limits and parameters the user configures in advance.

Daily Brief and Android Halo (34-36)

34. Daily Brief reads Gmail, Calendar, and Tasks each morning to generate a structured, prioritized summary of the day's commitments, outstanding items, and suggested next steps.

35. Daily Brief launches on May 19 and is immediately available for AI Plus, Pro, and Ultra subscribers in the United States.

36. Android Halo is a glowing on-screen indicator at the top of the Android display that shows the live status of background AI agents like Gemini Spark - available later in 2026.

Google Search Upgrades (37-43)

37. The intelligent Search box - the biggest redesign to Google Search in over 25 years - expands dynamically as the user types to accommodate longer, conversational queries.

38. The new Search box accepts multimodal inputs simultaneously - text, images, files, videos, and open Chrome browser tabs - reasoning across all of them in a single operation.

39. AI-powered query suggestions go beyond autocomplete by anticipating the user's underlying intent and proposing refined directions before the query is submitted.

40. AI Overviews and AI Mode merge into a single unified flow - users move seamlessly from standard results to AI Overviews to deeper AI Mode conversation without switching modes.

41. AI Mode has surpassed one billion monthly users and sees query volumes more than double every quarter since launch, reaching an all-time high for overall Search queries last quarter.

42. Information Agents in Search monitor topics around the clock - scanning news, blogs, social media, finance, shopping, and sports data to surface relevant changes proactively.

43. Antigravity-powered mini app building arrives in Search for AI Pro and Ultra subscribers during summer 2026, enabling lightweight custom dashboards and tools directly within the interface.

Universal Cart and Agentic Shopping (44-49)

44. Universal Cart is a persistent AI-powered shopping cart that follows users across Search, Gmail, YouTube, and the broader web simultaneously.

45. Universal Cart monitors price changes and deal availability across multiple retailers in real time, alerting users when meaningful savings opportunities appear.

46. Universal Cart identifies product incompatibilities before checkout - flagging mismatched components and suggesting alternatives automatically.

47. Universal Cart surfaces loyalty benefits and savings through Google Wallet integrations for supported merchants.

48. The Universal Commerce Protocol enables AI agents to browse retail inventories and act on behalf of users within the shopping ecosystem.

49. Universal Cart launches in the United States during summer 2026, with Gmail and YouTube integrations following in subsequent updates.

Ask YouTube (50-52)

50. Ask YouTube transforms video discovery - users ask a conversational question and jump directly to the most relevant segment of the most relevant video.

51. Ask YouTube supports contextual multi-turn follow-up questions, maintaining conversation continuity across an extended search session.

52. Ask YouTube is available from May 19 for YouTube Premium subscribers aged 18 and older in the United States at youtube.com/new, with a broader US rollout in summer 2026.

Google Workspace Updates (53-60)

53. AI Inbox brings new voice capabilities to Gmail - enabling dictated replies, new message composition, and inbox organization through spoken commands with Gemini context.

54. Google Pics treats every image element as a separately editable object, enabling precise modifications without regenerating the entire composition.

55. Google Pics generates posters, infographics, and marketing assets in a workflow comparable to professional graphic design software - directly within Workspace.

56. All Google Pics output carries automatic SynthID watermarking for content provenance and AI origin transparency.

57. Google Pics rolls out to AI Pro and Ultra subscribers during summer 2026, currently with trusted testers as of May 19.

58. Docs Live converts natural spoken language - including corrections and informal phrasing - into a finished, formatted Google Doc using Gemini context from Gmail, Drive, and Chat.

59. Docs Live rolls out globally in English for AI Pro and Ultra subscribers on Android and iOS during summer 2026, with similar capabilities coming to Gmail and Google Keep.

60. Stitch gains real-time collaborative design through a new Stitch Agent - multiple users co-design simultaneously with direct export to Antigravity and one-click Netlify publishing.

Google Flow and Creative Tools (61-66)

61. Google Flow is now available as a mobile app on Android in beta - with iOS coming soon - bringing AI filmmaking tools to on-the-go production workflows for the first time.

62. Flow Music is now available as a dedicated iOS app - with Android arriving shortly - enabling AI music creation from a mobile device.

63. Flow Music adds targeted lyric editing within specific sections of a song, enabling precise textual changes without regenerating the full composition.

64. Flow Music enables new song generation from a reference segment of an existing composition - building an original track anchored in a specified musical direction.

65. Gemini Omni integration in Google Flow brings multimodal video creation to the filmmaking platform, enabling cinematic AI video generation from any reference combination.

66. Google Flow supports AI agents for multi-action execution - agents can handle multiple simultaneous production steps within a single creative workflow.

Android XR Smart Glasses (67-71)

67. Android XR audio glasses - designed in partnership with Samsung, Qualcomm, Gentle Monster, and Warby Parker - launch in fall 2026 with both Android and iPhone compatibility.

68. Android XR audio glasses support real-time audio translation in the speaker's own voice, enabling live multilingual conversation without any additional device.

69. Android XR audio glasses provide Gemini voice interactions - the wearer receives AI responses through private audio audible only to them.

70. Android XR display glasses - demonstrated through the Xreal Project Aura reference design with a 70-degree field-of-view display and fingerprint sensor compute puck - follow at a later date.

71. Project Aura by Xreal features a redesigned compute puck with fingerprint sensor support and Gemini-powered XR integrations across Android apps from Google Play.

Antigravity 2.0 and Developer Infrastructure (72-80)

72. Antigravity 2.0 is Google's redesigned agent-first development platform running on Gemini 3.5 Flash, supporting parallel multi-agent workflows for simultaneous coding and testing.

73. Antigravity 2.0 includes a new command-line interface (CLI) giving developers terminal-based access to all agent capabilities without requiring a graphical interface.

74. Antigravity 2.0 launches a standalone desktop application bringing the full platform to local development environments with persistent project context.

75. The Antigravity SDK gives developers programmatic control over the agent harness for custom infrastructure deployments and specialized pipeline configurations.

76. Antigravity 2.0 now includes Google Cloud enterprise privacy protections - making the platform appropriate for regulated and enterprise development environments.

77. Antigravity 2.0 can generate an 8-bit game in approximately 8 seconds - demonstrated live during the keynote as a showcase of the platform's new generation speed.

78. Managed Agents in the Gemini API provision a fully operational AI agent with a remote sandbox through a single API call, removing manual orchestration infrastructure requirements.

79. WebMCP is a proposed open web standard that exposes structured JavaScript functions and HTML forms so browser-based AI agents can execute complex tasks more reliably - entering an experimental Chrome 149 origin trial.

80. Native Voice Support for Antigravity is available globally from May 19 - enabling voice-directed development workflows within the Antigravity platform.

SynthID and Content Credentials (81-85)

81. SynthID has now marked over 100 billion AI-generated images, videos, and audio files since its launch, establishing it as the world's most widely deployed AI content watermarking system.

82. SynthID verification has been used 50 million times globally through the Gemini app alone, confirming widespread adoption of AI content origin checking.

83. SynthID detection expands to Chrome through a right-click context menu on any image, allowing users to check AI origin on any webpage without switching tools.

84. SynthID and C2PA Content Credentials expand to Google Search - users can check whether any image in Search results was created by AI or is an unaltered camera original.

85. OpenAI, Kakao, and Eleven Labs confirmed adoption of SynthID-compatible content tagging standards - expanding cross-industry alignment on AI content provenance.

Gemini for Science (86-92)

86. Gemini for Science is a new suite of AI research tools combining deep reasoning, Deep Research capabilities, and experimental Google Labs tools for scientific exploration.

87. Science Skills connects Antigravity to over 30 major life science databases - including UniProt, AlphaFold Database, AlphaGenome API, and InterPro - from May 19 on GitHub and directly in Antigravity.

88. Gemini for Science enables complex bioinformatics and genomic analyses in minutes - workflows that previously required hours of manual expert labor.

89. Co-Scientist is a collaborative AI research partner within Gemini for Science, supporting hypothesis generation, literature synthesis, and experimental design.

90. The Paper Assistant Tool (PAT) is an experimental tool for agentic peer review and scientific paper validation, developed in partnership with major AI research conferences.

91. ScholarPeer is a companion experimental tool for collaborative scientific validation - developed through dedicated pilots with ICML, STOC, and NeurIPS.

92. Users can register interest in Gemini for Science experiments at labs.google/science, with gradual access opening from May 19 onward.

Project Genie, Pomelli, and Google Labs (93-96)

93. Project Genie integrates with nearly 20 years of Google Street View imagery - enabling users to generate interactive, navigable 3D environments anchored in real geographic locations.

94. Project Genie is available globally for Google AI Ultra subscribers from May 19, 2026.

95. Pomelli is a new Google Labs tool using AI agents to generate brand books, design guidelines, and functional websites from a high-level brand description in a single workflow.

96. Pomelli is part of the broader specialized agentic tools expansion across Google Labs, representing a new category of AI-first creative and brand infrastructure tooling.

Android XR, Smart Glasses Partners, and Hardware (97-98)

97. The first audio glasses - made in partnership with Gentle Monster, Warby Parker, and Samsung - arrive in fall 2026 and are compatible with both Android and iOS devices.

98. SynthID verification is now accessible by asking "Is this AI generated?" directly in Search via Lens, AI Mode, and Circle to Search on Android devices.

AI Subscription Pricing (99-100)

99. A new AI Ultra plan starts at $100 per month - replacing the $250 entry point - including 5x higher usage limits, 20 TB storage, Gemini Spark, Project Genie, and YouTube Premium. New and existing Ultra subscribers receive $100 in bonus Antigravity credits in a limited promotional window.

100. AI Pro at $19.99 per month now includes YouTube Premium Lite in select countries at no additional cost - and the previous $250 Ultra plan dropped to $200 with identical capabilities preserved.

New AI Models: Gemini 3.5 and Gemini Omni

Gemini 3.5 Flash: Frontier Intelligence with Action

The headline model at Google I/O 2026 is Gemini 3.5 Flash - the first in Google's latest model family, built to combine frontier-level intelligence with genuine agentic capability. Google processes more than three trillion tokens per day internally using this model, a figure that reflects both its deployment scale and its reliability under load. Gemini 3.5 Flash surpasses Gemini 3.1 Pro across every benchmark - coding accuracy, multimodal understanding, long-horizon agentic task completion, and real-world utility. It runs four times faster than comparable frontier models in terms of output tokens per second and delivers this performance at less than half the cost of competing frontier alternatives.

The model is generally available starting May 19 via the Gemini API in Google AI Studio, Android Studio, Antigravity 2.0, and across all major Gemini-powered consumer products. It powers AI Mode in Search as the new global default model from the same date. Gemini 3.5 Pro is already in internal use and will roll out to developers and subscribers within the following month. AI Mode, now running on Gemini 3.5 Flash, has surpassed one billion monthly users and continues to see query volumes double every quarter since its launch.

Gemini Omni Flash: Creation from Any Input

Google DeepMind CEO Demis Hassabis introduced Gemini Omni on stage at Google I/O 2026 - a new model series that represents what Google describes as a fundamental leap forward in world understanding, multimodal capability, and generative editing. Gemini Omni Flash, the first model in the series, accepts any combination of text, image, audio, and video as input and outputs AI-generated video grounded in real-world knowledge. It can simulate complex physical concepts including kinetic energy, gravity, and material behavior. Additionally, it translates dense scientific ideas into visually comprehensible video content.

Users can edit Omni-generated video through natural conversational prompts - applying cinematic zooms, swapping backgrounds, and changing characters without touching a traditional timeline interface. They can also create personal AI avatars that replicate their own voice and generate videos in their likeness. Every video produced by Gemini Omni carries an imperceptible SynthID digital watermark, allowing AI origin to be verified through the Gemini app, Gemini in Chrome, and Google Search.

Gemini Omni Flash is rolling out on May 19 for all Google AI Plus, Pro, and Ultra subscribers globally through the Gemini app and Google Flow. It is also available at no cost for users aged 18 and older through YouTube Shorts Remix and the YouTube Create App. Developer and enterprise API access arrives in the weeks following the keynote.

Gemini 3.5 and the Broader Model Roadmap

Beyond Flash, Google confirmed that future Omni model versions will support image and text outputs alongside video - eventually enabling any output type from any input combination. The Gemini 3.5 family will expand further with the Pro model, and the Gemini 3 series - including the existing Gemini 3 Pro Deep Think - continues to serve as the foundation for the most demanding enterprise and research applications.

New Custom Silicon: TPU 8t and TPU 8i

Supporting everything at Google I/O 2026 is a new generation of custom hardware. Google announced its eighth-generation Tensor Processing Units using a dual-chip approach for the first time - TPU 8t for large-scale pretraining and TPU 8i for inference workloads. TPU 8t delivers nearly three times the raw compute of the previous generation while providing up to two times better performance per watt. Both chips achieve the same efficiency improvement. Google can now distribute AI training seamlessly across more than one million TPUs globally using JAX and Pathways infrastructure - removing the previous constraint of single-datacenter training limits. This architecture creates the largest training cluster in the world and reduces the time required to train large, capable models from months to weeks.

The Gemini App: Redesign and New Capabilities

Neural Expressive: A New Design Language

The Gemini app received a complete visual overhaul built around a new design language called Neural Expressive. The updated interface introduces fluid animations, vibrant and responsive color palettes, haptic feedback integration, and revised typography throughout. The redesign is not cosmetic alone - it creates a more coherent structure for the app's rapidly growing feature set, with Gemini Omni, Gemini Spark, and Daily Brief each integrated into the new layout.

Compute-Based Usage Model

The Gemini app is transitioning from a fixed daily prompt limit to a compute-based measurement system. Each interaction now consumes resources proportional to the actual computational demands it places on the system. A short text query uses minimal compute. A complex video generation or long multi-step coding task consumes significantly more. Usage limits now refresh every five hours and count against a weekly compute budget - a more equitable system for users across different task types and usage patterns.

Regional Dialect Support for Voice

Voice output in the Gemini app now supports regional dialect selection. Users choose from localized speech variations within their primary language, receiving AI audio that reflects their specific regional linguistic context rather than a standardized neutral voice.

Personal Intelligence Expansion

Personal Intelligence - the feature that connects Gmail and Google Photos to AI Mode for more personalized answers - is expanding to nearly 200 countries and territories across 98 languages. This expansion makes Google's most contextually aware AI capabilities available at a genuinely global scale.

Gemini Spark: Your 24/7 Personal AI Agent

Gemini Spark is arguably the most consequential announcement at Google I/O 2026 for individual users. It is a 24/7 personal AI agent that runs entirely on Google Cloud infrastructure, keeping working on the user's behalf even when their phone is locked, their laptop is closed, or they are completely offline and unavailable.

Spark connects to Gmail, Google Docs, Sheets, and the broader Workspace ecosystem at launch. Third-party app support via MCP - the standard for connecting AI to external applications - arrives over the summer. The agent requires user approval before taking high-stakes actions, such as sending emails, completing purchases, or modifying documents. Lower-stakes monitoring, drafting, and information-gathering tasks run autonomously.

Practical capabilities demonstrated at the keynote include parsing credit card statements for overlooked recurring charges, monitoring inboxes and flagging deadlines from school or work communications, transcribing meeting notes into Google Docs and emailing them to attendees, and drafting project emails from a brief voice note. The Agents Payment Protocol embedded in Spark enables fully automated purchasing within spending limits the user configures.

Trusted testers received access during the week of the conference. Broader rollout to AI Ultra subscribers in the United States begins the following week. The significant cloud compute demands of running a continuous, personalized agent restrict initial access to the Ultra tier.

Daily Brief: Proactive AI Morning Summary

Daily Brief is a new proactive Gemini feature that reads Gmail, Calendar, and Tasks each morning to generate a structured, prioritized daily summary. It surfaces upcoming meetings, outstanding items, and context-aware suggested next steps. Daily Brief launched on May 19 and is available immediately for AI Plus, Pro, and Ultra subscribers in the United States.

Android Halo: Agent Activity at a Glance

Android Halo is a new on-screen status indicator that appears as a subtle glowing bar along the top edge of the Android phone display. It shows the live operational status of background AI agents - starting with Gemini Spark - so users maintain awareness of agent activity without opening any application. Android Halo is described as providing "subtle communication" that keeps users informed from any screen they are currently using. It arrives later in 2026, beginning with Gemini Spark integration.

Google Search: The Biggest Upgrade in 25 Years

Intelligent Search Box

Google I/O 2026 introduced the most significant redesign of Google's Search box in over 25 years. The new box expands dynamically as the user types, accommodating longer, more conversational queries that reflect how people actually formulate complex questions. It accepts text, images, files, videos, and open Chrome browser tabs as simultaneous inputs - all processed and reasoned across together in a single search operation. AI-powered query suggestions go beyond autocomplete, anticipating the user's underlying intent and proposing more refined search directions before the query is submitted.

Unified AI Search: AI Overviews Meet AI Mode

AI Overviews and AI Mode are now merged into a single, continuous AI Search flow. Users move fluidly from a standard results page into an AI Overview and then into deeper conversational AI Mode - all within the same interface without any mode-switching. This unified experience is live globally on desktop and mobile from May 19.

Information Agents in Search

A new class of persistent Information Agents operates within Search - monitoring topics and questions the user cares about, around the clock, across blogs, news sites, social platforms, financial data, shopping listings, and sports information. These agents surface meaningful changes proactively without the user needing to re-query. They land for AI Pro and Ultra subscribers during summer 2026, with mini app-building capabilities through Antigravity also arriving in the same window.

Universal Cart: Agentic Shopping Across Google

Universal Cart is Google's new shared, intelligent shopping cart that persists across Search, Gmail, YouTube, and the broader web. Items can be added from any Google surface. The cart monitors price changes across retailers, tracks deal history, alerts users when products return to stock, identifies product incompatibilities before checkout, and surfaces savings opportunities through Google Wallet integrations. Supporting merchants at launch include major names across electronics, apparel, home goods, and health and beauty.

The accompanying Universal Commerce Protocol enables AI agents to browse retail inventories. The Agents Payment Protocol in Gemini Spark allows fully automated checkout within user-defined spending parameters. Universal Cart launches in the United States during summer 2026, with Gmail and YouTube integrations to follow.

Ask YouTube: Conversational Video Search

Ask YouTube transforms how users discover information within YouTube. Users ask a conversational question and receive a structured answer that jumps directly to the most relevant segment within the most relevant video. The feature supports contextual follow-up questions across an ongoing conversation thread. It is available from May 19 for YouTube Premium subscribers aged 18 and older in the United States through youtube.com/new, with a broader US rollout planned for summer 2026.

Google Workspace: AI Inbox, Google Pics, Docs Live, and Stitch

AI Inbox

AI Inbox brings new voice capabilities to Gmail, enabling users to dictate replies, compose new messages, and organize their inbox through spoken commands. Gemini pulls context from across Gmail and Drive to produce contextually appropriate responses and summaries without requiring manual navigation.

Google Pics

Google Pics is a new AI image creation and editing tool built into Workspace. It treats every element within an image as a separately editable object - enabling precise modifications without regenerating the entire composition. Users can swap character expressions, change product colors, rewrite text elements within images, and build full posters, infographics, and marketing assets in a workflow comparable to professional design software. All output carries automatic SynthID watermarking. Google Pics is with trusted testers now and rolls out more broadly to AI Pro and Ultra subscribers during summer 2026.

Docs Live

Docs Live converts natural spoken language including false starts, mid-sentence corrections, and casual phrasing into a finished, formatted Google Doc using Gemini. The system pulls context from Gmail, Drive, and Google Chat to ensure the resulting document reflects accurate details and relevant information the user may not have mentioned explicitly. A similar voice-to-document capability is coming to Gmail and Google Keep. Docs Live rolls out in English globally to AI Pro and Ultra subscribers on Android and iOS during summer 2026.

Stitch: Real-Time Collaborative Design

Stitch - Google's AI-powered UI and website builder - received real-time collaborative design capabilities through a new Stitch Agent at Google I/O 2026. Multiple users can now co-design simultaneously within the same session. Stitch exports directly to Antigravity for development integration and supports one-click publishing to Netlify for immediate live deployment.

Google Flow, Flow Music, and Creative Tools

Google Flow is now available as a mobile app on Android in beta, with iOS coming soon. Flow Music - the companion AI music creation tool - is available as a dedicated mobile app on iOS, with Android arriving shortly. Flow Music adds targeted lyric editing within individual song sections and enables new song generation from a reference segment of an existing composition. Both apps bring Gemini Omni-powered creation capabilities to mobile production workflows for the first time.

Android XR Smart Glasses

Google I/O 2026 closed with the Android XR smart glasses announcement. Google, Samsung, and Qualcomm unveiled two variants. Audio glasses - designed with Gentle Monster and Warby Parker - launch in fall 2026 and deliver private Gemini audio responses audible only to the wearer. They support real-time audio translation in the speaker's own voice, navigation assistance, notification summaries, and Gemini voice interactions. Display glasses - demonstrated through the Xreal Project Aura reference design with a 70-degree field-of-view display and a redesigned compute puck with fingerprint sensor support - follow at a later date. Both versions are compatible with Android phones and iPhones. Pricing and regional availability details were not disclosed at the keynote.

Antigravity 2.0 and Developer Tools

Antigravity 2.0: Agent-First Development

Antigravity 2.0 is Google's upgraded, agent-first development platform. It runs on Gemini 3.5 Flash and supports parallel multi-agent workflows - where complex development tasks are distributed across specialized sub-agents that code and test simultaneously. A new command-line interface (CLI), a standalone desktop application, and the Antigravity SDK - giving developers programmatic control over the agent harness - all launched at Google I/O 2026. Enterprise-grade Google Cloud privacy protections are now built into the platform. Antigravity 2.0 can generate interactive 8-bit games in approximately 8 seconds, a capability demonstrated live during the keynote.

Managed Agents in the Gemini API

The Managed Agents capability allows developers to provision a fully operational AI agent with a remote sandbox through a single API call, eliminating the need to build and maintain underlying orchestration infrastructure manually.

WebMCP

WebMCP is a proposed open web standard allowing developers to expose structured JavaScript functions and HTML forms so browser-based AI agents can execute complex tasks more reliably. Its experimental origin trial begins in Chrome 149.

SynthID, Content Credentials, and AI Transparency

SynthID has now marked over 100 billion AI-generated images, videos, and audio files. Its verification system has been used 50 million times globally through the Gemini app alone. At Google I/O 2026, Google expanded SynthID detection to Chrome through a right-click menu on any image, to Google Search through Lens, AI Mode, and Circle to Search, and via the C2PA Content Credentials standard allowing users to ask "Is this AI-generated?" directly of any content they encounter. OpenAI, Kakao, and Eleven Labs all confirmed adoption of SynthID-compatible content tagging standards at the conference.

Gemini for Science

Gemini for Science is a new AI research suite connecting Antigravity to over 30 major life science databases including UniProt, AlphaFold Database, AlphaGenome API, and InterPro through Science Skills. This allows researchers to execute complex bioinformatics, structural biology, and genomic workflows in minutes rather than hours. A Co-Scientist collaborative AI research partner and experimental tools for agentic peer review including the Paper Assistant Tool (PAT) and ScholarPeer are also part of the suite. Science Skills are available on GitHub and directly within Antigravity from May 19. Users can register interest in experimental tools at labs.google/science.

Project Genie and Street View Worlds

Project Genie - Google DeepMind's interactive 3D world-generation model - now integrates with nearly 20 years of Google Street View imagery. Users can generate navigable, interactive 3D environments anchored in real-world geographic locations. Project Genie is available globally for AI Ultra subscribers.

Pomelli: Brand Building with AI Agents

Pomelli is a new Google Labs tool that uses AI agents to generate complete brand identities and functional websites from a high-level brand description. It produces brand books, design guidelines, and ready-to-publish web pages from a single starting description. Pomelli is part of the broader expansion of specialized agentic tools launched across Google Labs at Google I/O 2026.

New AI Subscription Tiers

Google restructured its AI plans significantly. A new AI Ultra plan starts at $100 per month - less than half the prior $250 entry point. It includes five times higher Gemini app usage limits than AI Pro, 20 TB of storage, priority Antigravity access, Gemini Spark access in the United States, Project Genie globally, and full YouTube Premium. The existing $250 plan dropped to $200 with identical capabilities. AI Pro remains at $19.99 per month and now includes YouTube Premium Lite for subscribers in select countries at no additional cost. New and existing AI Ultra subscribers receive a one-time $100 bonus in Antigravity credits during a limited promotional window.

What Google I/O 2026 Means for Your Career

The density and breadth of Google I/O 2026 confirms that AI is reshaping every professional domain simultaneously - from software development and data science to marketing, design, research, and commerce. Professionals who invest in structured AI expertise now will carry a compounding advantage over those who engage with these tools only at the surface level.

Those who pursue a Google Gemini Professional certification develop the platform-specific technical knowledge needed to build, evaluate, and deploy Gemini-powered applications across the full range of what Google I/O 2026 introduced from Gemini 3.5 Flash and Gemini Omni to Antigravity 2.0 and the Gemini API.

Furthermore, understanding the broader landscape of AI model behavior, agentic decision-making, safety considerations, and responsible deployment is now a baseline expectation in any role that intersects with AI at a professional level. An AI Certification provides this foundational grounding - enabling professionals to evaluate tools like Gemini Spark, Gemini Omni, and Managed Agents with structural technical understanding rather than surface awareness.

As AI tools become the operational infrastructure for every technology-adjacent role, professionals need verified credentials that demonstrate fluency across the broader technology ecosystem. A Tech Certification builds that recognized expertise across the domains that power everything announced at this year's conference from custom silicon and cloud infrastructure to agentic platforms and developer tools.

Finally, the commercial implications of Google I/O 2026 are most immediate for marketing and growth professionals. Universal Cart, Information Agents in Search, Ask YouTube, and AI-powered Workspace tools change how audiences discover products, engage with content, and convert across Google's ecosystem. A Marketing Certification equips professionals with the strategy framework to translate these shifts into measurable campaign performance, audience growth, and commercial outcomes.

FAQs

When did Google I/O 2026 take place?

Google I/O 2026 took place on May 19 and 20, 2026, at the Shoreline Amphitheatre in Mountain View, California, with a simultaneous global livestream on YouTube. The opening keynote was delivered by Google and Alphabet CEO Sundar Pichai.

What was the central theme of Google I/O 2026?

The event's central theme was the launch of the "agentic Gemini era" - a period in which AI transitions from answering questions to autonomously completing tasks across workflows, apps, devices, and the web on the user's behalf.

What is Gemini 3.5 Flash and what makes it significant?

Gemini 3.5 Flash is Google's newest foundation model, combining frontier intelligence with agentic task capabilities. It surpasses Gemini 3.1 Pro across all benchmarks, runs four times faster than competing frontier models, costs less than half the price of comparable alternatives, and powers every major Gemini product and API from May 19.

How many monthly users does AI Mode in Search have?

AI Mode has surpassed one billion monthly users as of Google I/O 2026 and continues to see query volumes more than double every quarter since launch. Search queries overall reached an all-time high in the most recent quarter.

What is Gemini Omni Flash?

Gemini Omni Flash is the first model in Google's new Omni series. It accepts any combination of text, image, audio, and video as input and outputs AI-generated video grounded in real-world knowledge. It supports conversational editing and includes SynthID watermarking on all generated content.

What can Gemini Spark do?

Gemini Spark can parse credit card statements, monitor email inboxes for deadlines, transcribe meeting notes into Google Docs, draft project emails from voice notes, and autonomously complete purchases within user-defined spending limits - all while running continuously in the cloud, independent of device state.

What is the new AI Ultra pricing?

A new AI Ultra plan starts at $100 per month, replacing the previous $250 entry point. The original $250 plan dropped to $200 with identical features. AI Pro remains $19.99 per month and now includes YouTube Premium Lite in select countries.

What are the TPU 8t and TPU 8i chips?

TPU 8t is Google's eighth-generation chip optimized for large-scale AI training, delivering nearly three times the raw compute of the previous generation. TPU 8i handles inference. Both chips deliver up to two times better performance per watt. Google can now distribute training across more than one million TPUs globally.

What is Android Halo?

Android Halo is a glowing status bar element at the top of the Android phone screen that shows the live activity of background AI agents like Gemini Spark. It provides at-a-glance agent awareness without requiring any app to be opened and arrives later in 2026.

What changed in Google Search at I/O 2026?

The Search box was completely redesigned for the first time in over 25 years - expanding dynamically, accepting multimodal inputs including Chrome tabs, and providing intent-based query suggestions. AI Overviews and AI Mode merged into a single unified flow. Persistent Information Agents that monitor topics around the clock are coming to Pro and Ultra subscribers in summer 2026.

What is the Universal Cart?

Universal Cart is a persistent AI-powered shopping cart that works across Search, Gmail, YouTube, and the web. It monitors prices, tracks deals, flags product incompatibilities, highlights savings through Google Wallet, and supports fully automated purchasing through the Agents Payment Protocol in Gemini Spark.

What is Ask YouTube?

Ask YouTube allows users to search YouTube conversationally and jump directly to the most relevant segment of the most relevant video. It supports multi-turn follow-up questions and is available now for US YouTube Premium subscribers aged 18 and older at youtube.com/new.

What is Docs Live?

Docs Live converts natural spoken language - including corrections and casual phrasing - into a finished, formatted Google Doc using Gemini. It pulls context from Gmail, Drive, and Chat to ensure accuracy. It arrives globally in English for AI Pro and Ultra subscribers on Android and iOS in summer 2026.

What are the Android XR smart glasses?

Two versions were announced: audio glasses designed with Gentle Monster and Warby Parker, launching fall 2026, and display glasses using the Xreal Project Aura reference hardware. Both work with Android phones and iPhones. They support real-time translation, navigation, notifications, and Gemini voice interactions.

What is Antigravity 2.0?

Antigravity 2.0 is Google's agent-first development platform, running on Gemini 3.5 Flash. It supports parallel multi-agent workflows, adds a new CLI, a desktop app, the Antigravity SDK, and Google Cloud enterprise privacy protections. It can generate an 8-bit game in approximately 8 seconds.

What is Google Pics?

Google Pics is a new Workspace image creation and editing tool that treats each image element as individually editable. It generates posters, infographics, and marketing assets, and watermarks all output with SynthID. It rolls out to AI Pro and Ultra subscribers in summer 2026.

What is Gemini for Science?

Gemini for Science connects Antigravity to over 30 major life science databases through Science Skills, enables complex bioinformatics and genomic workflows in minutes, and includes a Co-Scientist AI research partner and experimental peer review tools. Science Skills are available on GitHub and in Antigravity from May 19.

What SynthID milestones did Google announce at I/O 2026?

SynthID has watermarked over 100 billion AI-generated images, videos, and audio files. Its verification feature has been used 50 million times globally. Detection expanded to Chrome, Search, and Circle to Search. OpenAI, Kakao, and Eleven Labs all adopted SynthID-compatible content tagging.

What is Pomelli?

Pomelli is a new Google Labs tool that uses AI agents to generate brand books, design guidelines, and complete functional websites from a high-level brand description. It launched as part of the expanded specialized agentic tools ecosystem announced at Google I/O 2026.

What is Project Genie's new capability at Google I/O 2026?

Project Genie now integrates with nearly 20 years of Google Street View imagery, enabling users to generate navigable 3D environments anchored in real-world geographic locations. It is available globally for Google AI Ultra subscribers from May 19.