How did AI models achieve full marks at the IMO?

Question

Accepted Answer

AI Models Achieve Perfect Scores on International Mathematical Olympiad Questions

Artificial intelligence models have achieved a perfect score of 100 per cent on the International Mathematical Olympiad (IMO) questions, marking a significant advancement in AI's mathematical reasoning capabilities. Huawei's AI system, named "Celia," and Xiaohongshu's model, "dots-note-3.0," were announced to have attained this feat. These AI submissions were made under strict conditions, with human intervention prohibited, and the solutions were graded by the official IMO organizers.
AI Systems Tested Under Official Conditions

Tech companies were provided with the IMO problems only after the human contestants had completed their exams. The AI solutions were submitted within a specified time limit, mirroring the constraints faced by human participants. Crucially, any form of human intervention during the testing phase was strictly prohibited to ensure the integrity of the AI's performance. The submitted AI solutions were then forwarded to the IMO organizers for the official grading process.
Notable Performance from Other AI Models

The perfect scores achieved by Huawei and Xiaohongshu are part of a broader trend of AI excelling in complex mathematical challenges. At the 2024 IMO, Google's AI achieved a silver-medal score, successfully solving four out of the six problems presented. Furthermore, Deedy Das reported that four other advanced AI models—from OpenAI, Anthropic, Axiom Math, and Moonshot AI's "Kimi K3"—all achieved a perfect score of 42 out of 42 points on this year's IMO questions.
Previous AI Achievements in Mathematical Competitions

While this year marks the first instance of AI achieving a perfect score under the official IMO judging process, AI models have been steadily improving their performance in such contests. In 2025, AI models from companies like Google and OpenAI reached gold-level scores, demonstrating their ability to tackle advanced mathematical problems. However, they had not previously matched the perfect scores attained by top human contestants.

Reading history

Hide →

This history is cleared when the page is closed.

How did AI models achieve full marks at the IMO?

Loading

Was this article helpful?

Helpful

Not helpful

Report an issue

1

0

0

Share with friends