Google has announced the launch of Gemini 3.1 Pro, the latest iteration of its AI model set to power the Gemini app and NotebookLM. This new model is specifically engineered to tackle complex problem-solving and advanced reasoning challenges.
Alphabet CEO Sundar Pichai highlighted the enhanced capabilities of Gemini 3.1 Pro, emphasizing its proficiency in visualizing intricate concepts, consolidating data into a unified perspective, and facilitating creative endeavors.
Notably, Gemini 3.1 Pro surpasses its predecessors and competitors in various benchmark assessments. It achieved an impressive 77.1% score on ARC-AGI 2, showcasing superior logic and problem-solving abilities.
On the Humanity’s Last Exam benchmark, Gemini 3.1 Pro outperformed other models with a leading score of 44.4%. While Opus 4.6 briefly took the lead in certain scenarios, Gemini’s overall performance remains commendable.
According to Google Labs Vice President Josh Woodward, Gemini 3.1 Pro excels in agentic tasks, complex coding, and data synthesis projects, promising reduced errors, enhanced logic, and high-quality visual outputs.
As Gemini 3.1 Pro begins its rollout to consumers, the tech community anticipates the impact of this advanced AI model on problem-solving and reasoning applications.
Source: mint – technology