Google today launched Gemini 3, positioning it as its smartest AI model yet, and — in a first — made it immediately available in Google Search. According to Google, this integration means the model will power the “AI Mode” feature for paying users, offering conversational, generative responses instead of standard search links.
Executives at Google highlighted major gains in reasoning, coding, and multimodal capabilities. Gemini 3 also introduces a new assistant-like feature called Gemini Agent, which can automate complex tasks — like managing your inbox or booking travel — on behalf of users.
For developers, Google revealed Antigravity, a coding platform powered by Gemini 3 Pro that supports agentic workflows. This tool lets AI handle multi-phase coding tasks across IDE, terminal, and browser environments.
Google’s CEO, Sundar Pichai, called Gemini 3 “our most intelligent model,” and said embedding it in core products from day one underscores Google’s push to monetize AI while keeping it central to how people search, work, and build.
