In a landmark achievement, artificial intelligence has scored a perfect 100% at the International Mathematical Olympiad (IMO), previously a feat reserved for a select few human contestants. Two Chinese tech giants, Huawei and Xiaohongshu, announced that their AI models achieved full marks in this year's contest held in Shanghai.
The IMO, one of the world's most prestigious mathematics competitions, saw 666 contestants from various countries this year, with only seven human participants achieving a perfect score. Huawei's AI system, 'Celia,' and Xiaohongshu's model, 'dots-note-3.0,' were the first to announce their perfect scores, setting a new benchmark for AI in mathematical problem-solving.
Xiaohongshu highlighted that this is the first time a large language model has achieved a perfect score under the IMO's official judging process. Previously, models from tech firms like Google and OpenAI reached gold-level scores but fell short of the perfect mark achieved by human contestants in 2025.
The AI models were provided with the IMO problems after human contestants had taken the exam and were required to submit their solutions within a specified time limit. Huawei and Xiaohongshu both emphasized that no human intervention was allowed during testing, ensuring the integrity of the AI's performance.
This achievement underscores the rapid advancement of AI in complex problem-solving tasks. Deedy Das, a partner at Menlo Ventures, noted that several cutting-edge AI models from leading US and Chinese labs also scored perfectly on this year's IMO questions, indicating that AI capabilities have surpassed previous expectations.
