Google has announced the release of Gemini 3, its most intelligent AI model to date. According to Google, Gemini 3 is designed to help users bring any idea to life with enhanced reasoning and multimodal capabilities. This new model is now accessible across various Google products, including the Gemini app, AI Studio, and Vertex AI.
Gemini 3: Key Features and Capabilities
Gemini 3 Pro reportedly outperforms previous models in key benchmarks such as reasoning, multimodality, and coding. Google is also planning to release a "Deep Think" mode for Ultra subscribers, which promises to push the boundaries of AI intelligence even further for complex problem-solving.
Sundar Pichai, CEO of Google and Alphabet, highlighted the rapid adoption of Google's AI technologies. He noted that AI Overviews now have 2 billion users every month, the Gemini app surpasses 650 million users per month, and a significant portion of Cloud customers are using Google's AI solutions.
Generative UI: A New Approach to User Interfaces
In addition to Gemini 3, Google is also introducing Generative UI, a novel implementation that allows AI models to create immersive experiences and interactive tools on the fly. This technology is rolling out in the Gemini app and Google Search, starting with AI Mode.
Generative UI enables AI models to generate not just content, but an entire user experience, including web pages, games, tools, and applications. These interfaces are automatically designed and customized in response to any question, instruction, or prompt.
Dynamic View and AI Mode
The research on Generative UI is being implemented in the Gemini app through an experiment called dynamic view and in AI Mode in Google Search. This allows users to experience dynamic interfaces tailored to their needs, rather than selecting from a predefined catalog of applications.
According to Google, evaluations indicate that the interfaces from their Generative UI implementations are strongly preferred by human raters compared to standard LLM outputs.