Google’s new Gemini Pro model has recorded record-breaking benchmark scores.

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

Google’s new model is probably one of the most powerful LLMs out there. Onlookers have pointed out that Gemini 3.1 Pro appears to be a significant improvement over its predecessor, Gemini 3, which was already considered a superior AI tool when it launched in November.

On Thursday, Google also shared statistics from an independent benchmark called Humanity’s Last Exam, which showed it performed significantly better than its predecessor.

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

This launch comes as the AI ​​model war heats up and technology companies continue to release increasingly powerful LLMs designed for agent tasks and multi-level inference. Other major names, including OpenAI and Anthropic, have also recently released new models.