Skip to content

Benchmark

Frontiermath Tiers 1 3 V2

Math benchmarks measure how well models solve maths problems. Scores are percentages.

Models measured
93
Scores recorded
125
Sources
1
IQ weight
1.66

Score distribution

47.3%
Mean
45.3%
Median
28.4%
Std dev
9.5%
10th percentile
88.1%
90th percentile
0%
Min
0%95.1%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.66 for chat models, measured over 93 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

Showing 101–125 of 125 results.

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence)