Google has launched Gemini 3.1 Pro, a sophisticated AI model designed for advanced problem-solving in science and engineering. This upgrade achieves a remarkable 77.1% on the ARC-AGI-2 benchmark, significantly outperforming its predecessor. Available to developers via the Gemini API and Google AI Studio, it supports tasks like complex system synthesis and offers enhanced usage limits for Pro and Ultra plan users. Further enhancements are planned to optimize agentic workflows.