Google I/O 2026: AI Search Overhaul, Smart Glasses, and Subscription Shifts
The tech giant’s keynote centred on artificial intelligence, introducing autonomous digital assistants, real-time translation glasses, and new video generation tools.
Google I/O 2026 has commenced with a keynote that reaffirms the company’s strategic pivot towards artificial intelligence. The event marked the global rollout of an intelligent, AI-powered Search box designed to anticipate user intent rather than simply autocompleting text queries. This new interface accepts images, video files, and entire Chrome tabs as direct inputs, while the existing AI Mode, now powered by the newly unveiled Gemini 3.5 Flash, remains available for follow-up questions and corrections.
A significant addition to the ecosystem is Gemini Spark, a cloud-based digital assistant capable of autonomous task management. Spark can monitor credit card statements to identify hidden subscriptions, track updates from school emails, and interact with third-party applications such as OpenTable and Instacart. While the assistant can execute these tasks, Google stated it will require user confirmation before finalising any purchases or sending emails.
In the hardware sector, Google and Samsung teased a collaboration with optical partners Gentle Monster and Warby Parker, revealing first looks at two models of Android XR smart glasses. These devices feature real-time audio translation that mimics the speaker’s voice, text overlay capabilities, and photo capture functionality. Users will also be able to chat directly with Gemini while wearing the glasses.
Subscription structures for Google’s AI services have been adjusted to accommodate the new feature set. A new mid-range plan has been introduced at $100 per month, positioned between the standard $20 Pro plan and the top-tier Ultra option. The Ultra plan, discounted from $250 to $200, now offers 20 times the usage limits of the Pro plan, 20TB of cloud storage, and exclusive access to Project Genie, an experimental tool for building 3D worlds using Street View imagery, as well as the Antigravity coding tool.
Google also announced Gemini Omni Flash, a video generation model described as an evolution of previous iterations. The model can combine images, audio, video, and text inputs to generate high-quality videos grounded in real-world knowledge, with improved understanding of physical forces such as gravity and fluid dynamics. The announcement comes as competitors move into similar spaces, with Spotify expanding its personal podcast generation features and Meta launching Forum, a dedicated app for Facebook Groups conversations.
