Skip to content

Benchmark

Arena ELO Text Factuality - Exclude Ties

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

Models measured
169
Scores recorded
193
Sources
1
IQ weight
2.31

Score distribution

1,431.5
Mean
1,440
Median
52.6
Std dev
1,353.3
10th percentile
1,488
90th percentile
1,239.9
Min
1,239.91,545.9
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.31 for chat models, measured over 169 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 CarbonHigh· LMArena data1,545.9
2Gemini 4 NeonHigh· LMArena data1,529.2
3Claude Opus 5.5High· LMArena data1,519.7
4Claude Fable 5.1Max· LMArena data1,515.7
5Gemini 4 ArgonHigh· LMArena data1,508.5
6Gemini 4 HydrogenHigh· LMArena data1,506.7
7Muse WonderMax· LMArena data1,506.1
8Claude Opus 4.6High· LMArena data1,505.6
9Claude Opus 4.6· LMArena data1,501.5
10Claude Opus 5Max· LMArena data1,501.3
11MiMo V2.6 Pro· LMArena data1,498.9
12Gemini 3 Pro Preview· LMArena data1,494
13Claude Fable 5.2Max· LMArena data1,493.2
14Muse Spark 1.3Max· LMArena data1,492.6
15Claude Fable 5High· LMArena data1,491.8
16GPT 5.5High· LMArena data1,490.8
17GPT 5.4High· LMArena data1,490.5
18Muse Spark 1.2xHigh· LMArena data1,489
19Qwen3.7 Max Preview· LMArena data1,488.1
20Claude Opus 5High· LMArena data1,488
21Gemini 3.5 FlashHigh· LMArena data1,487.3
22Claude Opus 4.7High· LMArena data1,487.2
23Qwen3.5-Max-Preview· LMArena data1,486.7
24GPT 5.5· LMArena data1,486.7
25Gemini 3.7 FlashHigh· LMArena data1,485.5
26Claude Opus 4.7· LMArena data1,484.9
27Qwen3.8 Max· LMArena data1,483.9
28Kimi K3.1 ProMax· LMArena data1,482
29Gemini 3.8 FlashHigh· LMArena data1,479.9
30Gemini 3.6 FlashHigh· LMArena data1,479.9
31GPT-5.6 SolxHigh· LMArena data1,479.7
32Claude Sonnet 5.5xHigh· LMArena data1,479.4
33Gemini 3.1 Pro Preview· LMArena data1,479.4
34Gemini 3 Flash· LMArena data1,478
35Muse Spark· LMArena data1,477.4
36Kimi K3Max· LMArena data1,477.4
37GPT 5.4· LMArena data1,475.8
38GLM 5.3Max· LMArena data1,475.6
39Gemini 3.5 FlashMedium· LMArena data1,475
40Claude Opus 4.8High· LMArena data1,474.9
41Muse Spark 1.1· LMArena data1,473.2
42GPT-6.1 SolMax· LMArena data1,472.6
43DeepSeek V4.1 Flash 0910Max· LMArena data1,472.4
44GPT-5.6 TerraxHigh· LMArena data1,470.9
45GLM 5.2Max· LMArena data1,469.6
46MiMo V2.5 Pro· LMArena data1,469
47Claude Opus 4.8· LMArena data1,467.8
48GPT-6 AstraMax· LMArena data1,467.7
49GLM 5.3 Flash· LMArena data1,467.3
50Grok 4.5· LMArena data1,467

Showing 1–50 of 193 results.

Data credits: LMArena, CC BY 4.0 (licence)