Gemini DeepMind: Gold-Level Math Olympiad AI

Google's advanced Gemini Deep Think model has achieved gold-medal level performance in the 2025 International Mathematical Olympiad. It solved five of the six problems under competition-style timing. What's interesting is that both Google DeepMind and OpenAI reported gold-medal level scores. However, Google's result carries additional weight because the company worked with IMO organizers and had its solutions officially assessed. OpenAI's gold-level score did not go through the same official competition process. This distinction matters because the credibility of evaluation is crucial in a market filled with benchmark claims. Developers and buyers want to know if an AI system's reasoning can withstand external review. The bottom line is that AI models are now strong enough to compete near the top of demanding math contests, demonstrating impressive progress in formal reasoning.