Skip to content

Benchmark

Arena ELO Text Factuality - Industry Legal and Government

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

Models measured
116
Scores recorded
132
Sources
1
IQ weight
2.45

Score distribution

1,464.6
Mean
1,463.8
Median
32.6
Std dev
1,428
10th percentile
1,507
90th percentile
1,353.2
Min
1,353.21,526.1
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.45 for chat models, measured over 116 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 ArgonHigh· LMArena data1,526.1
2Claude Opus 5Max· LMArena data1,522.8
3Claude Opus 5.5High· LMArena data1,522.8
4Claude Opus 5High· LMArena data1,521.9
5GPT 5.5High· LMArena data1,519
6Claude Opus 4.6High· LMArena data1,517.4
7Muse WonderMax· LMArena data1,516.8
8Claude Fable 5.1Max· LMArena data1,515.4
9Claude Opus 4.6· LMArena data1,513
10GPT 5.4High· LMArena data1,510.5
11GPT-6.1 SolMax· LMArena data1,509
12Claude Opus 4.7High· LMArena data1,508.9
13Claude Fable 5High· LMArena data1,508.8
14Muse Spark 1.2xHigh· LMArena data1,507
15GPT-5.6 SolxHigh· LMArena data1,505.3
16Gemini 3 Pro Preview· LMArena data1,504.2
17GPT 5.5· LMArena data1,502.5
18Gemini 3.5 FlashHigh· LMArena data1,500.2
19GPT-6 AstraMax· LMArena data1,500.1
20Qwen3.5-Max-Preview· LMArena data1,498.4
21GPT 5.4· LMArena data1,498.2
22Claude Opus 4.7· LMArena data1,497.5
23Claude Sonnet 5.5xHigh· LMArena data1,497.5
24Gemini 3.7 FlashHigh· LMArena data1,497
25Kimi K3Max· LMArena data1,496.9
26Qwen3.8 Max· LMArena data1,495.6
27GPT-5.6 TerraxHigh· LMArena data1,495.6
28Claude Opus 4.8High· LMArena data1,495.3
29Gemini 3.8 FlashHigh· LMArena data1,494.8
30Gemini 3.1 Pro Preview· LMArena data1,494.6
31Muse Spark 1.3Max· LMArena data1,494.4
32Claude Opus 4.8· LMArena data1,493.4
33GLM 5.2Max· LMArena data1,492.4
34Gemini 3.5 FlashMedium· LMArena data1,490
35Grok 4.5· LMArena data1,486.4
36GPT-5.6 LunaxHigh· LMArena data1,486.2
37Gemini 2.5 Pro· LMArena data1,486.1
38Gemini 3.6 FlashHigh· LMArena data1,484.7
39GLM 5.3Max· LMArena data1,484.3
40Claude Opus 4.520251101· LMArena data1,482.2
41Muse Spark 1.1· LMArena data1,481.9
42MiMo V2.5 Pro· LMArena data1,481.1
43GPT 5.1High· LMArena data1,480.8
44Claude Sonnet 4.6· LMArena data1,480.2
45DeepSeek V4 ProHigh, 20260813· LMArena data1,480
46GLM 5.1· LMArena data1,477.8
47DeepSeek V4 Pro· LMArena data1,477.7
48GLM 5.3 Flash· LMArena data1,477
49Claude Opus 4.5High, 32K, 20251101· LMArena data1,476.9
50GPT 5.1· LMArena data1,476.8

Showing 1–50 of 132 results.

Data credits: LMArena, CC BY 4.0 (licence)