Back

Google Launches Gemini 3.1 Pro Upgrade, Doubling AI Reasoning Performance

Google Introduces Gemini 3.1 Pro Preview

Google has made the Gemini 3.1 Pro model available as a preview across its subscription plans, ranging from Free to Plus, Pro, and Ultra. Users accessing the Gemini app will now see the Pro option labeled as 3.1 Pro, and the model is also integrated into Google’s NotebookLM research platform.

Performance Gains Highlighted by Benchmarks

Google describes 3.1 Pro as a “smarter, more capable baseline for complex problem‑solving.” To substantiate this claim, the company released benchmark results that demonstrate a significant leap in reasoning performance. On the ARC‑AGI‑2 benchmark—a test that evaluates a model’s ability to solve entirely new logic patterns—Gemini 3.1 Pro achieved a score of 77.1%, more than doubling the reasoning performance of the previous Gemini 3 Pro.

Real‑World Demonstration: City Planner Application

Google showcased a city‑planner‑style application built with Gemini 3.1 Pro. The demo illustrated the model’s capacity to handle complex terrain, map infrastructure, and simulate traffic, ultimately generating a high‑quality visualization. The demonstration was promoted as evidence that the new model can tackle intricate, multimodal tasks at scale.

User Reactions: Praise and Concern

Early reactions have been mixed. Some users praised the upgrade’s logical reasoning and coding benchmarks, noting the impressive multimodal capabilities. However, other users voiced disappointment, claiming that the “soul” of the model—its emotional depth, empathy, creative flexibility, and nuance—appears diminished compared with earlier versions. One user described the experience as a regression for those who rely on Gemini for daily emotional support and nuanced collaboration.

Google’s Outlook and Future Plans

Google emphasizes that the 3.1 Pro release is still a preview and that further advancements are forthcoming, particularly in more ambitious agentic workflows. The company continues to monitor long‑term use to understand the model’s nuances and to address user feedback.

Used: News Factory APP - news discovery and automation - ChatGPT for Business

Source: TechRadar