Google has unveiled a new generation of artificial intelligence tools and platforms under its Gemini ecosystem, marking what the company describes as a major move into the “agentic era” of AI during the Google I/O 2026 developer conference in Mountain View, California.
The tech company announced a wide range of AI products, infrastructure upgrades and consumer-focused features aimed at bringing autonomous AI agents into everyday digital experiences.
Speaking at the conference, Google CEO Sundar Pichai said the company is now prioritising the rollout of AI agents for ordinary users after years of developing the technology for businesses and software developers.
“Gemini 3.5 and Antigravity are unlocking a new world of agents and agentic capabilities. We’ve been bringing agents to developers and enterprises for a while. Now we are super focused on bringing the power of agents, safely and securely, to consumers so that it works for everyone,” said Pichai.
Google said the latest developments are part of its long-term strategy to become an AI-first company, adding that the journey began more than a decade ago.
According to the company, its ecosystem now combines custom-built hardware, AI research models, cloud systems and consumer applications to support innovation across multiple sectors.
The company revealed that usage of its AI platforms has grown sharply over the last two years. Google said the number of tokens processed monthly by its AI systems has increased from 9.7 trillion two years ago to more than 3.2 quadrillion today.
Google also said more than 8.5 million developers are building experiences using its AI models every month, while customers using Google Cloud processed over one trillion tokens each during the past year.
The company further highlighted growing adoption of Gemini-powered consumer products.
Google said AI Overviews in Search now serves 2.5 billion monthly active users, while AI Mode crossed one billion monthly users within a year.
The Gemini application has also surpassed 900 million monthly active users, with users generating more than 50 billion images through Nano Banana AI models integrated into the platform.
Among the new features announced is Ask YouTube, an AI-powered video discovery tool designed to help users locate important sections in long videos by responding to detailed questions.
Google said testing for the feature begins immediately ahead of a wider rollout in the United States later this year.
The company also introduced Docs Live, a voice-powered feature that allows users to create and edit documents through spoken commands instead of typing.
Google said the feature will be rolled out to subscribers during the summer, with similar voice features expected to be added to Gmail and Keep later on.
To support the growing demand for AI services, Google announced plans to sharply increase spending on infrastructure.
The company said annual capital expenditure is projected to reach about $190 billion, equivalent to nearly Sh24 trillion, this year, compared to $31 billion, or about Sh4 trillion, in 2022.
Google said much of the investment will go toward custom-built silicon chips, including the launch of eighth-generation Tensor Processing Units named TPU 8t and TPU 8i.
The company also unveiled Gemini 3.5 Flash, a new AI model that it says delivers frontier-level intelligence while operating at much faster speeds.
According to Google, the model performs better than Gemini 3.1 Pro across several benchmarks while running four times faster than competing frontier AI systems.
Google said the model could also help large organisations lower the cost of running AI-powered services.
The company additionally announced Antigravity 2.0, a desktop platform designed to manage groups of autonomous AI agents, alongside Gemini Spark, a personal AI assistant capable of handling background tasks and integrating with services such as Chrome, email and chat platforms.
Google said it is also adding agentic capabilities into Search through Information Agents that can continuously collect information and perform tasks for users automatically.
Other tools previewed during the conference included Daily Brief, Google Flow, Google Pics, Gemini-powered smart glasses and Gemini for Science, an experimental AI research system focused on speeding up discoveries in life sciences.