Google Introduces New Agentic Era With Gemini

At its annual developer conference, Google unveiled a broad set of artificial intelligence updates that signal the company’s deeper push into what it describes as the “agentic” era of AI, where systems can complete tasks, make decisions, and operate with greater autonomy.

Speaking at Google I/O 2026, CEO Sundar Pichai said the company is now focused on bringing AI agents to mainstream consumers after earlier deployments aimed at developers and enterprises.

“Gemini 3.5 and Antigravity are unlocking a new world of agents and agentic capabilities,” Pichai said during the keynote address.

The announcements span Google’s AI models, search products, infrastructure, developer platforms, and experimental consumer tools. Much of the focus was placed on expanding the capabilities of Gemini, the company’s flagship AI platform, while integrating conversational AI features more deeply across Google products.

Google said usage of its AI systems has grown significantly over the past two years. According to the company, its models now process more than 3.2 quadrillion tokens per month, up from 480 trillion a year ago. The company also reported that more than 8.5 million developers are now building with Google AI models monthly, while APIs process about 19 billion tokens every minute.

The company linked much of this growth to Gemini integrations across its ecosystem. Google said AI Overviews in Search now reach 2.5 billion monthly users, while AI Mode has crossed one billion monthly active users within a year of launch. The Gemini app itself has reportedly surpassed 900 million monthly active users.

Several new AI-powered product features were also introduced.

Among them was Ask YouTube, a conversational video discovery feature that allows users to search within videos and jump directly to relevant sections. The feature enters testing immediately, with a wider U.S. rollout planned later this year.

Google also introduced Docs Live, a voice-driven feature for document creation and editing. The tool allows users to generate and revise documents using spoken prompts rather than typed instructions. Similar voice capabilities are expected to expand into Gmail and Keep.

The company additionally previewed updates to Maps, where conversational interactions powered by Gemini are now being integrated through a feature called Ask Maps.

A major portion of the event focused on infrastructure and computing power required to support AI at scale. Google said it expects to spend approximately $190 billion this year, largely directed toward AI infrastructure and custom silicon development.

The company introduced two new versions of its Tensor Processing Units (TPUs), its in-house AI chips designed for model training and inference. Google said the eighth-generation chips improve performance efficiency while supporting large-scale AI workloads distributed across multiple data centers.

Google also announced Gemini 3.5 Flash, a new AI model focused on faster response times and lower operational costs. According to the company, the model is designed for coding, reasoning, and other real-world tasks while operating more efficiently than previous versions.

The model is now available through Google products and APIs, while Gemini 3.5 Pro is expected to launch next month.

Alongside the models, Google expanded Antigravity, its platform for coordinating AI agents. The updated Antigravity 2.0 platform is designed to manage multiple autonomous AI systems and serve as a central environment for developers building agent-based applications.

The company also previewed several AI agent initiatives aimed at consumers and enterprise users.

One of them, Gemini Spark, is designed to operate continuously in the background, handling long-running tasks and connecting with tools across Google services. Another initiative, described as “agentic search,” would allow AI systems to gather information and complete actions on behalf of users over extended periods.

Additional experimental products showcased at the event included AI-powered smart glasses, collaborative workspaces, creative editing tools, and a research-focused platform intended to support scientific discovery using AI models connected to life science databases.

The announcements come as competition intensifies among major technology companies racing to commercialize generative AI and autonomous systems. Companies including Microsoft, OpenAI, Meta and Amazon have all accelerated investments in AI infrastructure, large language models, and AI assistants over the past year.

Google’s latest announcements indicate the company is positioning Gemini not only as a chatbot or productivity assistant, but as the foundation for a broader ecosystem of autonomous AI systems embedded across consumer and enterprise products.