Google's annual I/O event once again served as a platform for the tech giant to unveil a suite of groundbreaking updates and additional features across its product ecosystem, with artificial intelligence (AI) taking center stage. Prior to the main event, Google had already introduced numerous AI enhancements to its Android 17 software, including innovative AI-generated widgets. The conference further highlighted new Gemini models and expanded AI integrations within Google Search and Workspace, emphasizing a deepened commitment to intelligent systems that enhance user experience and productivity.

A significant announcement from Google's CEO, Sundar Pichai, was the introduction of Gemini 3.5 Flash. This latest iteration of the Gemini model is touted for combining "frontier intelligence with action," offering unparalleled speed and performance. According to Koray Kavukcuoglu, CTO of Google DeepMind and Chief AI Architect, Gemini 3.5 Flash rivals large flagship models on multiple dimensions. It boasts speeds up to four times faster in output tokens per second compared to other frontier models and demonstrates superior performance in challenging coding and agentic benchmarks. Specifically, it excels in Terminal-Bench 2.1 (76.2%), GDPval-AA (1656 Elo), and MCP Atlas (83.6%), alongside leading multimodal understanding (84.2% on CharXiv Reasoning). This formidable balance of speed and performance positions Gemini 3.5 Flash as an ideal tool for complex, long-horizon agentic tasks, drastically reducing development and auditing times, often at less than half the cost of competing models. Its enhanced agentic coding capabilities are now integrated into Search to deliver more intelligent experiences, and it has been made the default model for the Gemini app and AI Mode in Search globally. Gemini 3.5 Flash is generally available through Google Antigravity, the Gemini API (in Google AI Studio and Android Studio), Gemini Enterprise Agent Platform, and Gemini Enterprise.

Accompanying Gemini 3.5 Flash is Gemini Omni, a powerful new capability designed to allow users to create any output from virtually any input, starting with video. Building on last year's Nano Banana model, which aided in photo restoration and design, Gemini Omni aims to unlock a new world of agents and agentic capabilities for consumers. It offers an intuitive way to edit video using natural language, where each instruction builds coherently on the last, ensuring character consistency, realistic physics, and scene memory for endless continuity. Google describes Omni as combining an intuitive understanding of physics with Gemini’s vast knowledge of history, science, and cultural context, bridging the gap from photorealism to meaningful storytelling. While it can convert various references (image, text, video, audio) into a single, cohesive output, initial support for audio inputs will begin with voice references, with other audio types to follow.

Google also heralded a new era for AI Search, noting that AI Mode has rapidly surpassed one billion monthly users within a year of its debut, with queries more than doubling each quarter. This surge indicates that users are leveraging Search's expanded capabilities more than ever before. At I/O, Google announced its next strategic move: merging the best attributes of a traditional search engine with the cutting-edge power of AI. Liz Reid, Google’s VP and Head of Search, revealed that advanced model capabilities are being integrated into Search with new AI features, enabling users to interact with agents simply by asking questions. A significant overhaul, described as its biggest upgrade in over 25 years, introduces a new, intelligent AI-powered Search box. This upgraded Search, powered by Gemini 3.5 Flash, is more intuitive, providing more space for users to describe their exact needs. It can now anticipate user intent, formulating questions with AI-powered suggestions beyond mere autocomplete. Users can input queries using text, images, files, videos, or even Chrome tabs. Furthermore, the ability to ask follow-up questions directly from an AI Overview and engage in a conversational back-and-forth with AI Mode significantly enhances the search experience. As an added benefit, Google is introducing new agentic capabilities within Search to assist users with shopping. Importantly, Google is expanding Personal Intelligence in AI Mode to nearly 200 countries and territories across 98 languages, making it accessible without a subscription.

The Gemini app experience has been completely redesigned, featuring a new aesthetic language termed "Neural Expressive." This updated interface boasts fluid animations, vibrant colors, overhauled fonts, and haptic feedback. A live conversational experience is now deeply integrated, allowing seamless switching between typing quick questions and engaging in free-flowing voice conversations. Google also plans to offer regional dialects, enabling users to choose a voice that resonates with them. During the event, Google provided a preview of Android Halo, an upcoming feature designed to offer at-a-glance visibility into an agent's current tasks. Android Halo will display subtle communications at the top of the phone screen, indicating the agent's status when it takes on a task, enters live mode, or sends a message, ensuring users can monitor progress without interrupting their current activity. Android Halo is slated for release later this year.

For the more than four billion users who rely on Google Workspace apps like Gmail, Docs, and Drive, Google unveiled several compelling updates. New voice capabilities will empower users to brainstorm, organize thoughts, and complete tasks in Gmail, Docs, and Keep using only their voice. An all-new app, Google Pics, reimagines image generation and editing, offering ultimate precision for projects ranging from professional to everyday creative endeavors. Built on the latest Nano Banana model, Google Pics simplifies complex image creation, from designing event flyers to tweaking digital illustrations. AI Inbox is expanding to Google AI Plus and Pro subscribers, bringing new updates for enhanced inbox management. A significant new addition is Gemini Spark, a 24/7 personal AI agent within the Gemini app, designed to help users navigate digital life, take action on their behalf under their direction, and integrate seamlessly with Workspace applications. Furthermore, Gmail Live enables instant inbox searching via voice commands, such as "What’s my flight’s gate number?" or "What’s going on at my kid’s school this week?" Similarly, Docs Live acts as a thought partner and co-writer, helping users accelerate first drafts using voice. It organizes thoughts, structures documents, and, with user permission, pulls relevant details from Gmail, Drive, Chat, and the web, facilitating outlining, tone refinement, and seamless creation. In Keep, users can now simply "brain dump" ideas, and the app will not only understand their rambles but also automatically transform them into organized notes and lists.

Finally, Google elaborated on the evolution of Google Flow, its AI creative studio introduced last year for filmmakers in over 140 countries, offering video and image generation and editing capabilities. With the integration of Gemini Omni, Google Flow will now allow creatives to blend real-world inspiration with generated content and iterate conversationally. Omni Flash will further enhance character consistency, ensuring identity and voice are preserved across every scene. Google is also expanding Flow's accessibility with the launch of dedicated mobile apps: the Google Flow app (currently in beta for Android, with iOS coming soon) and the Google Flow Music app (available now on iOS, with Android coming soon). The Flow app is available for users aged 18 and older.