AI vs the world's best teen mathematicians
Year: 2026
Type: report
Abstract: Full interactive study, all charts and the latest version: https://iqmetrics.org/research/ai-vs-math-olympiad/ At the 2025 International Mathematical Olympiad, the only AI result the IMO itself confirmed would have placed joint 27th among 630 of the world's best teenage mathematicians. According to IQ Metrics' analysis of the official results, Google DeepMind's Gemini Deep Think scored 35 of 42, the lowest gold score: 26 humans scored more and 46 scored the same. Every AI that year missed Problem 6, which 6 contestants solved in full. In 2026, two companies said their models scored 42, a total 7 humans also reached; the IMO has not published a statement on them. Data: imo-official.org and the labs' own reports, October 2026. Where every AI result at the International Mathematical Olympiad would have ranked among the human contestants, 2024-2026, the problems AI could not solve, and 67 years of Olympiad results: medal cut-offs, girls, top countries and the medallists who went on to win the Fields Medal. Key findings: Gemini Deep Think: joint 27th of 630. Its IMO-certified 35 of 42 at the 2025 Olympiad equalled the lowest gold score. 26 contestants scored more, 46 scored the same and 558 (88.6%) scored less. The problem no AI could do. In 2025 Gemini, OpenAI's model, ByteDance's Seed Prover and Harmonic's Aristotle all failed Problem 6. Only 6 of 630 human contestants (1%) solved it in full. AI's blind spots are not human ones. In 2024 AlphaProof solved Problem 6, which only 5 of 609 humans solved, but missed Problem 5, which 154 humans solved. Its 28 points, with up to three days per problem, ranked joint 59th of 609: one point short of gold. 2026: claims of 42 out of 42. Xiaohongshu and Huawei said their models scored full marks, and an informal test gave four more models 42. Only 7 of 666 human contestants did that. The IMO has published no statement on the AI entries. Olympiad medallists become top mathematicians. 16 of the 38 Fields Medallists since 1990 (42.1%) won an IMO medal at school, 13 of them gold, including two of the four 2026 medallists. Girls are 1 in 9 contestants, as in 1959. 76 of the 666 contestants in 2026 were girls (11.4%), the same share as at the first Olympiad (6 of 52, 11.5%). The low point was the 1970s, at 2%. Files: the report (PDF), the data (ai-vs-math-olympiad-2026-iqmetrics.csv) and a README with the column list. Method: https://iqmetrics.org/research/methods/ Licence: CC BY 4.0. Credit "IQ Metrics Research, AI vs the world's best teen mathematicians 2026" with a link to https://iqmetrics.org/research/ai-vs-math-olympiad/
Authors: Falkner, Leon, Bhosale, Mahesh
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7224253461