AI Achieves Perfect Score at the International Mathematical Olympiad

International Mathematical Olympiad history was made this year after artificial intelligence achieved a perfect score for the first time at the world’s most prestigious mathematics competition.

Chinese technology companies Huawei and Xiaohongshu announced that their AI models each earned 100 per cent by correctly solving every problem from the 2026 International Mathematical Olympiad (IMO), matching the performance of the competition’s highest-scoring human contestants.

According to both companies, their AI systems completed the examination under official IMO evaluation conditions without any human assistance during testing.

Xiaohongshu said its AI model, dots-note-3.0, became the first large language model to achieve full marks through the competition’s official judging process. Huawei also announced that its AI system, Celia, demonstrated advanced problem-solving abilities across multiple areas of mathematics.

The companies received the IMO questions only after human contestants had completed the examination. They submitted their AI-generated solutions within the official time limit, with organisers grading the answers under the same standards applied to student competitors.

This year’s International Mathematical Olympiad took place in Shanghai, where 666 students from around the world competed. Only seven human participants achieved perfect scores, highlighting the exceptional difficulty of the contest.

The achievement marks another major milestone in artificial intelligence. Last year, AI systems reached silver-medal level, while models developed by companies including Google and OpenAI achieved gold-level performances in 2025 without attaining full marks.

Industry experts say the rapid improvement reflects how quickly AI reasoning abilities have advanced. Venture capital investor Deedy Das revealed that he independently tested this year’s IMO problems on four leading AI models developed by OpenAI, Anthropic, Axiom Math and Moonshot AI. According to Das, all four systems achieved the maximum score of 42 points.

The latest results suggest that advanced AI models are now capable of solving some of the world’s most challenging mathematical problems at a level equal to the best young mathematicians, underscoring the rapid progress of artificial intelligence in complex reasoning and scientific problem-solving.