Google presented its annual Google I/O conference yesterday to showcase a wide set of new AI tools and updates across its product ecosystem. The company placed a strong focus on new Gemini models, AI agents, and deeper integration across Search, Android, and creative apps. The announcements also highlighted growing competition with Apple ahead of its upcoming developer conference.
Google I/O works as Google’s main developer event, similar in role to the Apple Worldwide Developers Conference. This year, Google positioned AI as the core of almost every product experience. The updates also give a preview of how future search, creation, and productivity tools may evolve.
Gemini Becomes the Center of Google’s AI Push
Google introduced several updates under its Gemini umbrella, with new models designed for faster performance and richer media generation.
Gemini Omni is the most advanced version shown. It is designed to generate content from multiple types of input. It can handle text, images, and video together. It also improves the simulation of physics concepts like gravity and motion. The model combines Gemini intelligence with generative systems like Nano Banana and Veo.
Users can also edit videos through conversational prompts. They can upload clips and modify elements directly inside them. Google said Omni starts with video, but it may expand to all output types in the future. Gemini Omni Flash is the first release of this Omni family. It is available today inside the Gemini app for supported users.
Gemini 3.5 Flash is another major release. Google says it blends strong reasoning with fast response times. It performs better than 3.1 Pro across many benchmarks. The company also claims it matches top-tier models while staying faster. It is now available across Google products and APIs. Gemini 3.5 Pro is still in testing inside Google. It is expected to launch next month.
A Redesigned Gemini App and New AI Agents
The Gemini app has also been redesigned with a new “Neural Expressive” style. The update brings smoother animations, richer colors, and improved typography. It also adds haptic feedback on mobile devices. The redesign is rolling out across desktop, Android, and iOS today.
Google also added regional dialect support, which will arrive in the coming months. Paid users on Plus, Pro, and Ultra plans can already access Gemini Omni inside the app for video creation and editing. AI agents are now part of Gemini as well. A new Daily Brief agent can prepare personalized summaries for users each day. This feature is rolling out today for paid subscribers.
On Mac, Gemini is becoming more tightly integrated with the system. Users can select files in Finder and trigger actions using the function key and voice commands. In a demo, Google showed how Gemini could create and send an email using files stored locally and Gmail in Chrome. Voice features and Gemini Spark support are expected later this year.
Google also introduced Gemini for Science, which includes tools for research and collaboration. A Co-Scientist feature is designed to support researchers in long-term projects.
AI Content Detection Becomes Part of Chrome and Gemini
Google is expanding content transparency using C2PA credentials. This system helps identify whether images are captured by a camera or generated by AI. It can also detect AI edits in real photos.
Users will be able to right-click images inside Google Chrome and ask Gemini to analyze whether AI was used. This move pushes more transparency into online media as generative tools grow.
Google also updated its developer-focused tool, Antigravity 2.0. The platform is designed for agent-based coding workflows and competes with tools like Copilot-style assistants.
Antigravity 2.0 runs on Gemini 3.5 Flash and is optimized for speed and token efficiency. Google claims it is up to 12 times faster inside coding workflows. The tool is available globally starting today.
Gemini Spark Introduces Always-on AI Agents
A major highlight was Gemini Spark, a personal AI agent that runs in the cloud. It operates continuously on virtual machines through Google Cloud, meaning it does not require a laptop to stay active.
Spark can run tasks in the background and handle multi-step workflows. It can plan tasks, break them into steps, and execute them over time. It is accessible through the Gemini app, email, or messaging systems.
The system also supports integration with Google tools today. Third-party app support through MCP is coming soon. Spark will first launch for Google AI Ultra subscribers in the United States next week. Chrome support is planned for later this summer.
Google Search Becomes More AI-Driven
Google Search is also undergoing a major transformation. AI Overviews and AI Mode now run on Gemini 3.5 Flash. Google says Search is now an AI-first product. The search box has been redesigned to accept more than text. Users can input images, videos, files, and even browser tabs. Google described it as the biggest change to search in more than 25 years.
The company is also merging AI Overviews and AI Mode into a single interface. This creates a more unified experience across search results. Search will also support AI agents that can track live updates. These agents can monitor stock movements, website changes, and product drops. Users can receive alerts when changes happen online.
AI Expands into Coding, Shopping, and Dynamic Results
Search is also gaining agentic coding features. It can generate interactive tools, dashboards, and widgets directly in results. These features use Antigravity and Gemini 3.5 Flash. Google said this system will roll out this summer.
Shopping is also becoming more automated. A Universal Cart will let users collect products across websites, apps, and services. Users can later check out either through Google or third-party retailers.
Google is also introducing a Universal Commerce Protocol and an Agent Payment Protocol. These systems allow AI agents to handle purchases within user-defined limits. This feature will later connect with Gemini Spark.
YouTube and Docs Gain Conversational AI Tools
YouTube is getting a new Ask YouTube feature. It lets users ask questions about videos and get contextual answers. The feature supports follow-ups and is currently in testing in the United States.
Google Docs is also getting a new tool called Docs Live. Users can speak or type rough ideas, and Gemini will turn them into full documents. Google described it as a way to turn a “brain dump” into structured writing quickly.
Hardware Moves into AI-powered Glasses
Google also revealed Android XR audio glasses. These glasses provide always-on access to Gemini with private audio responses.
They can take photos, handle calls, play music, and connect to apps. The design comes from partnerships with Gentle Monster and Warby Parker, while Samsung supports hardware development. The glasses will work with both Android and iOS devices and launch this fall.
Pricing Changes and New Creative Tools
Google introduced a new AI Ultra plan priced at $100. The top-tier Ultra plan has also been reduced from $250 to $200. On Android, a new system called Android Halo will show AI agent activity at the top of the screen later this year.
Creative tools are expanding; Google Photos will allow users to design posters, flyers, and infographics inside Workspace. It also adds SynthID watermarking for AI-generated content.
A new tool called “Stitch” brings app and web design features similar to Figma’s. It will support real-time collaboration and direct exports to Antigravity and Netlify. Google Flow, a creative AI studio, is also gaining Gemini Omni support and new agent-based workflows.
Many of these features are already rolling out today, while others will arrive later this year. Google is clearly pushing toward a fully AI-driven ecosystem across search, productivity, coding, and hardware. The announcements set the stage for a direct comparison with Apple’s next big software reveal at WWDC in June.
