Problem Statement

OpenAI o3

Score: 7/7

Gemini 2.5 Pro Deep Think

Score: 7/7

Frequently Asked Questions

Which AI models solved IMO 2025?

The primary AI models that achieved gold-medal standards on the IMO 2025 problems are Google DeepMind's Gemini 2.5 Pro Deep Think and OpenAI's o3 reasoning model.

How did OpenAI o3 perform on IMO 2025?

OpenAI o3 solved difficult problems like Problem 1, 2, and Geometry tasks with high accuracy, demonstrating advanced chain-of-thought reasoning capabilities comparable to top human competitors.

Did Google DeepMind Gemini 2.5 Pro Deep Think win a gold medal?

Yes, the DeepMind system achieved a score equivalent to a silver/gold medal threshold, solving 5 out of 6 problems officially, marking a historic milestone in AI mathematical reasoning.

Where can I find the official proof steps?

This website provides the full, side-by-side official proof steps extracted from the technical reports of both OpenAI and Google DeepMind for IMO 2025.

Official Source Links: