At its annual developer conference in Mountain View, California, Google unveiled a sweeping set of artificial intelligence announcements at Google I/O 2026, signaling what executives described as the company’s transition into the “agentic era” of AI.
The company introduced a new generation of Gemini models, upgraded AI infrastructure, and a suite of autonomous AI agents designed to handle complex, multi-step tasks for consumers and enterprises alike.
“We’ve been bringing agents to developers and enterprises for a while,” said Sundar Pichai, chief executive officer of Google. “Now we are super focused on bringing the power of agents, safely and securely, to consumers so that it works for everyone.”
Pichai said the strategy builds on Google’s decade-long transformation into an AI-first company, combining investments in custom silicon, cloud infrastructure, research, and consumer products. He described artificial intelligence as “the most profound way” to advance Google’s mission of improving lives at global scale.
Google said adoption of its AI ecosystem has accelerated dramatically over the past two years. The company now processes more than 3.2 quadrillion tokens per month, up from 480 trillion a year ago and 9.7 trillion two years ago. More than 8.5 million developers use Google AI models each month, while the company’s APIs handle roughly 19 billion tokens every minute.
The company also highlighted strong momentum across its consumer products, driven largely by Gemini integrations. Google said 13 of its products now have more than one billion users each, with five surpassing three billion users globally.
Among the company’s flagship AI offerings, AI Overviews in Search now reaches 2.5 billion monthly active users, while AI Mode has surpassed one billion monthly users within a year. Google said the Gemini app has grown to more than 900 million monthly active users, with daily requests increasing more than sevenfold over the past year.
Google also showcased several new AI-powered consumer experiences aimed at making interactions more conversational and personalized.
One of the new features, Ask YouTube, is designed to transform video discovery by surfacing relevant moments within videos based on nuanced user questions. The feature enters testing immediately and is scheduled for a broader U.S. rollout this summer.
Another feature, Docs Live, enables users to generate and edit documents entirely through voice conversations. Powered by advances in Google’s audio models, the tool allows users to verbally dictate ideas and have them transformed into formatted documents. Voice capabilities are also expected to expand to Gmail and Keep later this year.
To support the rapid expansion of AI services, Google announced major infrastructure investments. The company expects to spend approximately $190 billion this year, up from $31 billion in annual capital expenditures in 2022, with much of the investment focused on custom AI chips.
Google unveiled its eighth generation of Tensor Processing Units, TPU 8t and TPU 8i. The chips are optimized separately for training and inference workloads and deliver up to twice the performance-per-watt of previous generations.
The company also introduced Gemini 3.5 Flash, a new AI model focused on balancing frontier-level intelligence with high-speed performance. Google said the model outperforms Gemini 3.1 Pro on several coding and real-world task benchmarks while operating four times faster than competing frontier models.
Gemini 3.5 Flash is now available across Google products and APIs, while Gemini 3.5 Pro is expected to launch next month.
Alongside the new model, Google launched Antigravity 2.0, a desktop platform designed to manage groups of autonomous AI agents. The platform runs a specialized version of Gemini 3.5 Flash optimized for rapid execution.
The company also previewed Gemini Spark, a persistent personal AI agent capable of operating continuously in the background through Google Cloud virtual machines. Google said Spark can connect to external tools, manage long-term workflows, and integrate with products including Gmail, Chrome, and chat applications. Beta testing for U.S. subscribers to Google AI Ultra is scheduled to begin next week.
Additional announcements included Daily Brief, a personalized AI-generated morning digest; Google Flow, a collaborative AI workspace for brainstorming and creative production; Google Pics, a new AI-powered image editing platform; and Gemini-powered intelligent eyewear designed for hands-free assistance and contextual information tracking.
Google also introduced Gemini for Science, an experimental research platform aimed at accelerating life sciences discoveries by connecting AI systems to more than 30 scientific databases.