Google has launched Gemini 2.0, its most advanced artificial intelligence model to date, capable of reasoning, multitasking, and acting on users’ behalf under supervision. The new model introduces “Deep Research,” a feature designed to assist with complex topics by generating comprehensive reports. Starting Wednesday, it will be available in Gemini Advanced.
Sundar Pichai, CEO of Google and Alphabet, emphasized the company’s commitment to developing agentic AI models. “Gemini 2.0 is built for the agentic era, enabling smarter interactions with multimodality, native tool use, and advanced reasoning. This brings us closer to a universal assistant,” Pichai said during the announcement.
Currently accessible to developers and trusted testers, Gemini 2.0 is expected to integrate with Google’s Search and the broader Gemini platform. Google also announced the experimental Gemini 2.0 Flash model, set to roll out to all Gemini users, expanding access to its cutting-edge capabilities.
Gemini 2.0 builds on its predecessor by offering advanced features for AI Overviews, now reaching over 1 billion users. The model can handle complex tasks such as solving advanced math equations, processing multimodal queries, and even coding. According to Pichai, “If Gemini 1.0 was about organizing and understanding information, Gemini 2.0 is about making it actionable.”
The release comes on the heels of Google’s announcement of Willow, its latest quantum chip. Willow significantly reduces error rates, representing a leap forward in AI and quantum computing capabilities. Together, these innovations reflect Google’s broader vision of redefining AI’s role in daily tasks and research.
