// benchmark explainer
AIME, explained
By the TrueIQ editorial team at Ahzy Capital · Last reviewed Oct 15, 2026 · How we work
The American Invitational Mathematics Examination is a real US maths competition. AI labs use each year's problems as a test of multistep maths reasoning.
| Run by | Mathematical Association of America (the competition) |
|---|---|
| Format | Each answer is an integer from 0 to 999 |
| Size | 30 problems a year (AIME I and II, 15 each) |
| Common versions | AIME 2024, AIME 2025 |
What it measures
Problems need several steps of algebra, geometry, number theory or combinatorics. Because answers are single integers, grading is automatic and exact.
Where it falls short
- Only 30 problems a year, so one problem is worth over 3 percentage points.
- Older years' problems and solutions are all over the internet, so newer years are more reliable.
- Labs often report the average of many attempts, which isn't the same as one try.
How to read a score
Look for which year was used and how many attempts were averaged. Gaps of a few points are within noise.
Where to see current scores
We don't republish scores. Leaderboards change weekly, and numbers depend on how the test was run. Check these instead:
- Artificial Analysis: AIME results (independently run)
- MAA: AMC/AIME competitions