Google’s new Gemini Pro model has record benchmark scores — again

Thursday Google issued the latest version of Gemini Pro, its powerful LLM. The model, 3.1, is currently available in preview and will be released generally soon, the company said.

Google’s new model may be one of the most powerful LLMs yet. Onlookers have noted that Gemini 3.1 Pro appears to be a big step up from its predecessor, Gemini 3, which was already considered a very capable AI tool upon its release in November.

On Thursday, Google also shared stats from independent benchmarks — such as one called Humanity’s Last Exam — that showed it significantly outperformed the previous version.

Gemini 3.1 Pro also received praise from Brendan Foody, the CEO of AI startup Mercor, whose benchmarking system, APEX, is designed to measure how well new AI models perform real-world professional tasks. “Gemini 3.1 Pro now tops the APEX Agents leaderboard,” Foody said in a social media postadding that the model’s impressive results show “how quickly agents are progressing in real knowledge work.”

The release comes as the AI model wars are heating upand tech companies continue to release increasingly powerful LLMs designed for agentic work and multi-step reasoning. Other big names – including OpenAI and Anthropic – have also released new models recently.