OpenAI's reasoning system achieved a gold medal-level performance at the 2025 International Olympiad in Informatics (IOI), ranking sixth overall among 330 human participants and first among AI competitors. The AI system improved from the 49th percentile to the 98th percentile within a year, without any IOI-specific training, and operated under the same constraints as human contestants, including a five-hour time limit and no internet access. This accomplishment follows recent gold medal-level achievements by OpenAI in other competitions such as the International Mathematical Olympiad (IMO) and AtCoder. The system was developed by a core team of three people over a focused two-month sprint, utilizing general-purpose reinforcement learning and scaled test-time compute to solve complex verification tasks. OpenAI's success in these prestigious programming competitions highlights rapid advancements in AI reasoning capabilities beyond the GPT-5 model.