News Summary

OpenAI has reached a significant milestone in AI development with its experimental model achieving gold medal-level performance at the 2025 International Math Olympiad (IMO). The model solved five out of six complex algebra and pre-calculus problems, earning 35 out of 42 points, a feat that only 67 of the 630 contestants managed to achieve. This accomplishment highlights the model’s capability in handling intricate mathematical problems with human-like reasoning, a longstanding challenge in AI. Despite this achievement, OpenAI has no immediate plans to release this model to the public, and the upcoming GPT-5 will not include this advanced mathematical capability.