By Kenrick Cai
MOUNTAIN VIEW, California, May 19 (Reuters) – Alphabet’s Google on Tuesday put AI agents directly into its search box and rolled out a faster, cheaper version of its Gemini model, aiming to blunt gains by rivals Anthropic and OpenAI among enterprise customers.
Alphabet also reinforced its core consumer products including Search and YouTube by announcing a set of agents that will autonomously complete tasks like making purchases and monitoring real estate listings in real time.
The moves, unveiled at Google’s annual I/O developer conference in Mountain View, California, mark its first major showcase since last winter’s update to its Gemini AI model that helped it regain ground in a high-stakes AI race and underline Google’s push to turn its vast consumer reach into an edge in AI.
“I think it’s been an incredible year, all of the relentless shifting, the rapid advances in technology. It’s been a period of hyper progress,” Pichai said.
Pichai revealed new products including Docs Live, which can help a user draft a document through a real-time voice conversation, and Ask YouTube, a conversational search interface.
“We are super-focused on bringing the power of agents safely and securely to consumers, so that they work for everyone,” Pichai said. Some of those tools run on Gemini 3.5 Flash, a model designed for coding and automated tasks.
Pichai said it matches leading AI systems, runs four times faster and is twice as cheap as competing models.
Pichai also said that Google’s Gemini AI chatbot now has 900 million monthly users – more than doubling in a year. Its AI Overviews feature in Search now has 2.5 billion monthly users, while AI Mode has about 1 billion, he added.
“When people use our AI-powered features in search, they use search more,” Pichai said.
Google parent Alphabet recently came within striking distance of Nvidia as the world’s most valuable company. It is looking to fortify its standing on Tuesday with a refreshed version of Gemini, plus new products and features built with the model’s capabilities.
ENTERPRISE SHIFT
Nick Fox, a senior vice president in charge of Google’s most lucrative business unit which includes Search and Ads, said direct access to agents is the “biggest reinvention of the search box in 25 years.”
Google also introduced Gemini Spark, a new agent that pulls information from apps like Gmail and Drive to draft reports and manage schedules.
The company lowered the cost of its top-tier AI Ultra subscription plan, which gives users access to higher AI usage limits and advanced AI models, to $200 a month from $250. It also announced a $100-a-month version of the Ultra subscription it said was tailored for developers and other work-related users.
As competition shifts to big-spending business customers, Google is stressing lower costs.
“We’ve heard that many companies are already blowing through their annual token budgets, and it’s only May,” Pichai said, referring to the units of data processed by AI models.
He added that heavy users such as big companies could save more than $1 billion a year by switching to Google’s models, which he said offer similar performance at a fraction of the cost.
GOOGLE LAUNCHES GEMINI OMNI VIDEO TOOL
The company also unveiled Gemini Omni, a new video model that Google executives sought to cast as a successor to the Nano Banana image generator, which attracted 13 million first-time users in just four days in September in what has marked one of Google’s few viral AI moments. Omni can generate videos with customizable audio and text.
Demis Hassabis, who runs Google’s DeepMind AI lab, said Gemini Omni represented the next step in Google’s vision to create a “world model,” which can simulate the physical nature of the world.
“Starting with video, but over time, Omni will be able to generate any output from any input,” Hassabis said.
Google has pushed to turn its vast consumer reach into an edge in AI, connecting Gemini to personalized user data across its suite of products that includes Chrome, Gmail, and YouTube.
Search was Alphabet’s biggest revenue driver in 2025, when it reported $402.8 billion in total revenue. The company is ramping up spending on AI infrastructure, expecting $180 billion to $190 billion in capital expenditures this year.
AD REVENUE GROWTH EASES INVESTOR CONCERNS
Advertising revenue, including from search ads, has continued to burnish Google’s growth in recent quarters, assuaging investor fears that AI could disrupt Google’s products and soften its market dominance.
AI rivals OpenAI and Anthropic have been gearing up for IPOs, focused on capturing lucrative enterprise customers. Google is also expected to dedicate time at the conference to focus on businesses, particularly software developers who have been a primary source of enterprise AI revenue.
In 2025, the company hired key staff from popular AI code generation startup Windsurf in a $2.4 billion deal to bolster its efforts around coding assistant Antigravity, which competes with Anthropic’s market-leading Claude Code software development tool.
Google is increasingly positioning digital assistants it calls agents, which can do complicated tasks autonomously, as the linchpin of its strategy to monetize AI. Pichai and other executives spoke about this at a business-oriented cloud computing conference last month.
Google Cloud CEO Thomas Kurian said at the time the company was saving the bulk of its coding-related announcements for the I/O conference.
(Reporting by Kenrick Cai in Mountain View, California, and Deborah Mary Sophia in Bengaluru; Editing by Sayantani Ghosh, David Gregorio and Matthew Lewis)
Comments