One of the largest reboots to Google’s product family in company history occurred at the company’s annual developer conference this year, Google I/O 2026, putting AI in everything from Search, Gmail and Shopping to the company’s hardware, research, and more. The newest Gemini models launched there, the Gemini Omni model is deemed by the company as “a huge leap forward for multimodality that can produce content based on all kinds of input, even video,” and along with that Gemini 3.5
Flash is a faster, more cost-efficient model that’s said to outshine many larger rivals on a number of important coding and analytics benchmarks. The Google I/O address by CEO Sundar Pichai cited that “AI Overviews are now being used by over 2.5 billion monthly active users” while “Gemini has now exceeded 900 million monthly active users,” up more than double from a year ago. Google also announced the progression to what it referred to as “agentic AI”, systems that, rather than just responding to questions, act on behalf of the user. Its newest model, Gemini Spark is planned to be a personal AI agent that runs on a user’s mobile phone or laptop constantly in the background, acting on behalf of the user. Google also launched Universal Cart, an AI-enabled shopping tool that manages purchases from all retailers in one place. On the hardware front Google showed off its new AI-powered audio glasses developed with partners including Samsung, Gentle Monster, and Warby Parker which put Gemini’s voice capabilities inside regular glasses, as well as the construction of the company’s largest AI training cluster ever. The company sees I/O 2026 as marking a new phase for the company, one where “AI isn’t simply a feature built on top of its services and products, but rather the foundation”.
