Despite groundbreaking performances from AI giants Google and OpenAI, human brilliance still reigned supreme at the 2025 International Mathematical Olympiad (IMO) in Queensland, Australia.
For the first time ever, AI models reached gold medal-level scores — both Google’s Gemini and OpenAI’s reasoning model scored 35 out of 42. Yet, five human contestants achieved perfect scores, reminding the world that natural intelligence still has the edge!
Google proudly confirmed its model solved 5 of the 6 brutal math problems — a major leap from its silver-level showing in 2024. OpenAI also impressed with equally high marks, saying all answers were independently graded by past IMO medalists.
While AI solved problems faster this year, IMO officials couldn’t verify how much computing power was used — or if there was human help behind the scenes.
Still, it’s a huge leap forward for AI in advanced reasoning and math… but human minds remain undefeated
