Skip to content

Benchmark

Arena ELO - Industry Mathematical

People compare model answers and choose which they prefer. Higher scores are better.

Models measured
348
Scores recorded
402
Sources
1
IQ weight
1.25

Score distribution

1,321.4
Mean
1,367.7
Median
154.1
Std dev
1,094.5
10th percentile
1,481.4
90th percentile
833.6
Min
833.61,566.4
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.25 for chat models, measured over 348 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 data1,566.4
2Gemini 4 ArgonHigh· LMArena data1,553.8
3Claude Opus 5High· LMArena data1,543.3
4Claude Fable 5.1Max· LMArena data1,543.3
5Claude Opus 5Max· LMArena data1,539
6Gemini 4 HydrogenHigh· LMArena data1,537.2
7Gemini 4 CarbonHigh· LMArena data1,537.1
8Claude Opus 5.5High· LMArena data1,536.6
9Claude Opus 4.6High· LMArena data1,525
10Gemini 3.8 FlashHigh· LMArena data1,518.9
11Claude Opus 4.6· LMArena data1,516.3
12Claude Fable 5High· LMArena data1,516
13MiMo V2.6 Pro· LMArena data1,514.4
14Claude Sonnet 5.5xHigh· LMArena data1,509.1
15Muse WonderMax· LMArena data1,508.5
16Claude Opus 4.7High· LMArena data1,508.2
17Kimi K3Max· LMArena data1,506.5
18GLM 5.3Max· LMArena data1,505
19Claude Fable 5.2Max· LMArena data1,503.9
20Gemini 3.7 FlashHigh· LMArena data1,500.9
21GLM 5.3 Flash· LMArena data1,496.8
22GPT 5.4High· LMArena data1,496.6
23Gemini 3.5 FlashMedium· LMArena data1,496
24Qwen3.8 Max· LMArena data1,495.3
25GPT-6.1 SolMax· LMArena data1,495
26Claude Opus 4.8High· LMArena data1,495
27Claude Opus 4.7· LMArena data1,494.3
28GPT 5.5High· LMArena data1,494.1
29GPT-5.6 SolxHigh· LMArena data1,492.9
30MiMo V2.5 Pro· LMArena data1,492
31Gemini 3.5 FlashHigh· LMArena data1,491.5
32Muse Spark 1.3Max· LMArena data1,491.5
33ERNIE 5.1· LMArena data1,491.3
34Qwen3.5-Max-Preview· LMArena data1,491.1
35GLM 5.2Max· LMArena data1,490
36Gemini 3.6 FlashHigh· LMArena data1,489.2
37GPT 5.5· LMArena data1,488.5
38Muse Spark 1.1· LMArena data1,483
39MiMo V2.6 Flash· LMArena data1,481.9
40Claude Sonnet 4.6· LMArena data1,481.6
41Claude Opus 4.8· LMArena data1,481.4
42Kimi K2.6· LMArena data1,480.3
43DeepSeek V4.1 Flash 0910Max· LMArena data1,479.5
44Grok 4.5· LMArena data1,479.5
45Claude Sonnet 5High· LMArena data1,478.9
46GPT-6 AstraMax· LMArena data1,477.2
47Gemini 3.1 Pro Preview· LMArena data1,477
48GPT-6.1 TerraMax· LMArena data1,476.6
49Hy3 (Deep Infra)· LMArena data1,476.3
50GPT-5.6 TerraxHigh· LMArena data1,475.8

Showing 1–50 of 402 results.

Data credits: LMArena, CC BY 4.0 (licence)