Skip to content

Benchmark

Arena ELO - Math

Math benchmarks measure how well models solve maths problems. Higher scores are better.

Models measured
350
Scores recorded
405
Sources
1
IQ weight
1.24

Score distribution

1,316
Mean
1,367.4
Median
152.6
Std dev
1,089
10th percentile
1,474.8
90th percentile
837.8
Min
837.81,547.6
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.24 for chat models, measured over 350 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Gemini 4 NeonHigh· LMArena data1,547.6
2Gemini 4 ArgonHigh· LMArena data1,542.9
3Claude Fable 5.1Max· LMArena data1,532.8
4Claude Opus 5High· LMArena data1,532
5Claude Opus 5Max· LMArena data1,531.7
6Gemini 3.8 FlashHigh· LMArena data1,525.5
7Claude Opus 5.5High· LMArena data1,522.8
8Claude Opus 4.6High· LMArena data1,517.5
9Claude Fable 5High· LMArena data1,516.5
10Gemini 4 HydrogenHigh· LMArena data1,515.1
11Claude Opus 4.6· LMArena data1,509.8
12GLM 5.3 Flash· LMArena data1,507.7
13Gemini 3.7 FlashHigh· LMArena data1,506.3
14Gemini 3.5 FlashHigh· LMArena data1,505.4
15Gemini 3.6 FlashHigh· LMArena data1,505
16Qwen3.8 Max· LMArena data1,504.7
17GLM 5.3Max· LMArena data1,501.9
18Claude Opus 4.7High· LMArena data1,498.5
19Kimi K3Max· LMArena data1,496
20MiMo V2.6 Pro· LMArena data1,493.9
21Muse Spark 1.3Max· LMArena data1,493.8
22Claude Sonnet 5.5xHigh· LMArena data1,493.8
23Muse WonderMax· LMArena data1,492.8
24Claude Fable 5.2Max· LMArena data1,490.7
25Claude Opus 4.7· LMArena data1,488.6
26GPT 5.4High· LMArena data1,487.6
27Gemini 3.5 FlashMedium· LMArena data1,487.4
28GPT 5.5· LMArena data1,485.9
29Claude Opus 4.8High· LMArena data1,485.3
30Gemini 3.1 Pro Preview· LMArena data1,485.1
31Qwen3.7 Max Preview· LMArena data1,485.1
32GLM 5.2Max· LMArena data1,484.2
33ERNIE 5.1· LMArena data1,482.2
34Muse Spark 1.1· LMArena data1,481.5
35Inkling FP4· LMArena data1,481.2
36MiMo V2.5 Pro· LMArena data1,480.1
37DeepSeek V4.1 Flash 0910Max· LMArena data1,479.7
38GPT 5.5High· LMArena data1,476.3
39GPT-6.1 SolMax· LMArena data1,475
40Gemini 3 Pro Preview· LMArena data1,475
41Qwen3.5-Max-Preview· LMArena data1,474.8
42Kimi K2.6· LMArena data1,474.6
43Hy3 (Deep Infra)· LMArena data1,473.9
44Muse Spark 1.2xHigh· LMArena data1,473.5
45Gemini 3 Flash· LMArena data1,473
46GLM 5.1· LMArena data1,472
47GPT-5.6 SolxHigh· LMArena data1,471.4
48Gemma 4 26B A4B· LMArena data1,470.5
49Claude Opus 4.8· LMArena data1,469.6
50Kimi K2.5Thinking· LMArena data1,469

Showing 1–50 of 405 results.

Data credits: LMArena, CC BY 4.0 (licence)