Skip to content

Benchmark

Arena ELO Text Factuality - Industry Mathematical

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

Models measured
102
Scores recorded
117
Sources
1
IQ weight
2.25

Score distribution

1,453.4
Mean
1,452.4
Median
30.2
Std dev
1,416.5
10th percentile
1,489.5
90th percentile
1,363.5
Min
1,363.51,520.2
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.25 for chat models, measured over 102 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Claude Opus 5High· LMArena data1,520.2
2Gemini 4 ArgonHigh· LMArena data1,513.1
3Claude Opus 5Max· LMArena data1,510.6
4Claude Fable 5.1Max· LMArena data1,506.3
5Claude Opus 5.5High· LMArena data1,505.4
6Claude Opus 4.6High· LMArena data1,503.8
7Claude Opus 4.6· LMArena data1,499.4
8GPT 5.4High· LMArena data1,497.9
9Claude Fable 5High· LMArena data1,496.6
10Gemini 3.8 FlashHigh· LMArena data1,496.3
11GPT 5.5High· LMArena data1,492
12GPT-6 AstraMax· LMArena data1,489.8
13Muse Spark 1.3Max· LMArena data1,489.5
14GPT-5.6 SolxHigh· LMArena data1,488.9
15Claude Opus 4.7High· LMArena data1,488.3
16Kimi K3Max· LMArena data1,487.8
17ERNIE 5.1· LMArena data1,485.8
18Gemini 3.5 FlashMedium· LMArena data1,484.7
19GPT 5.5· LMArena data1,484.7
20Claude Opus 4.8High· LMArena data1,483.8
21Grok 4.5· LMArena data1,483
22GPT-5.6 TerraxHigh· LMArena data1,482.9
23Gemini 3.7 FlashHigh· LMArena data1,482.6
24Claude Opus 4.7· LMArena data1,482.5
25Claude Opus 4.8· LMArena data1,481.9
26GLM 5.3 Flash· LMArena data1,481.8
27MiMo V2.5 Pro· LMArena data1,481.5
28Gemini 3.6 FlashHigh· LMArena data1,480.7
29GLM 5.3Max· LMArena data1,480.1
30Kimi K2.6· LMArena data1,476.7
31Qwen3.8 Max· LMArena data1,476.5
32Claude Sonnet 5High· LMArena data1,475.2
33GPT 5.4· LMArena data1,474.9
34GLM 5.2Max· LMArena data1,474
35Claude Sonnet 4.6· LMArena data1,473.5
36Gemini 3 Pro Preview· LMArena data1,471.9
37GPT-5.6 LunaxHigh· LMArena data1,471.4
38Claude Opus 4.520251101· LMArena data1,469.5
39Gemini 3.5 FlashHigh· LMArena data1,469.5
40Inkling FP4· LMArena data1,468.9
41Gemini 3.1 Pro Preview· LMArena data1,467.2
42Qwen3.7 Plus· LMArena data1,466.9
43Hy3 (Deep Infra)· LMArena data1,465.9
44Grok 4.6High· LMArena data1,465.8
45MiMo V2.6 Flash· LMArena data1,463.6
46Muse Spark 1.1· LMArena data1,463.2
47Muse Spark 1.2xHigh· LMArena data1,463.1
48Claude Opus 4.5High, 32K, 20251101· LMArena data1,461.1
49Claude Sonnet 4.5High, 32K, 20250929· LMArena data1,460.8
50Kimi K2.5Thinking· LMArena data1,460.2

Showing 1–50 of 117 results.

Data credits: LMArena, CC BY 4.0 (licence)