Google has launched Gemini 3.1 Pro, an upgraded AI model that boasts a more than twofold boost in reasoning performance, now surpassing competitors in the AI field. The development comes in response to rapid advancements made by rivals such as OpenAI and Anthropic, which had previously overtaken Google after its short-lived lead with Gemini 3 Pro. The new model excels in rigorous logic benchmarks, achieving a verified score of 77.1% on ARC-AGI-2, designed to assess a model’s capability to tackle unfamiliar logic patterns. Early reviews from enterprise partners like JetBrains and Databricks have indicated significant improvements in reliability and efficiency, showcasing the model’s potential for complex coding and data reasoning tasks.
Google boosts AI model with Gemini 3.1 Pro, doubling reasoning performance
