Skip to content

Benchmark

Arena ELO Text Factuality

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

Models measured
170
Scores recorded
194
Sources
1
IQ weight
2.26

Score distribution

1,435.9
Mean
1,441.9
Median
39.3
Std dev
1,379.4
10th percentile
1,478.9
90th percentile
1,299.6
Min
1,299.61,522.2
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.26 for chat models, measured over 170 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 data1,522.2
2Gemini 4 NeonHigh· LMArena data1,509.3
3Claude Opus 5.5High· LMArena data1,504.8
4Claude Fable 5.1Max· LMArena data1,498.7
5Muse WonderMax· LMArena data1,493.7
6Gemini 4 HydrogenHigh· LMArena data1,493.6
7Claude Opus 4.6High· LMArena data1,492.7
8Gemini 4 ArgonHigh· LMArena data1,492.3
9Claude Opus 4.6· LMArena data1,488.2
10Claude Opus 5Max· LMArena data1,487.8
11MiMo V2.6 Pro· LMArena data1,485.8
12GPT 5.5High· LMArena data1,485.2
13GPT 5.4High· LMArena data1,484.4
14Claude Fable 5High· LMArena data1,483.3
15Claude Fable 5.2Max· LMArena data1,483.1
16Gemini 3 Pro Preview· LMArena data1,481.5
17GPT 5.5· LMArena data1,481.4
18Muse Spark 1.3Max· LMArena data1,481.4
19Qwen3.7 Max Preview· LMArena data1,479.7
20Muse Spark 1.2xHigh· LMArena data1,478.9
21Kimi K3.1 ProMax· LMArena data1,478.8
22Gemini 3.5 FlashHigh· LMArena data1,478.3
23Claude Opus 4.7High· LMArena data1,478.1
24Claude Opus 5High· LMArena data1,477.6
25GPT-5.6 SolxHigh· LMArena data1,476.2
26Claude Opus 4.7· LMArena data1,476.1
27Gemini 3.7 FlashHigh· LMArena data1,476.1
28Qwen3.5-Max-Preview· LMArena data1,475.4
29GPT 5.4· LMArena data1,474.9
30Qwen3.8 Max· LMArena data1,473.3
31Claude Sonnet 5.5xHigh· LMArena data1,473.3
32Gemini 3.6 FlashHigh· LMArena data1,472.1
33Gemini 3.1 Pro Preview· LMArena data1,471.6
34Kimi K3Max· LMArena data1,470.5
35Claude Opus 4.8High· LMArena data1,470.3
36GPT-6.1 SolMax· LMArena data1,469.8
37GPT-5.6 TerraxHigh· LMArena data1,469.4
38Gemini 3.8 FlashHigh· LMArena data1,469.1
39GLM 5.3Max· LMArena data1,468.7
40Muse Spark· LMArena data1,468.1
41Gemini 3 Flash· LMArena data1,468
42GPT-6 AstraMax· LMArena data1,467.9
43DeepSeek V4.1 Flash 0910Max· LMArena data1,467.6
44Gemini 3.5 FlashMedium· LMArena data1,467.1
45GPT-6 LunaMax· LMArena data1,466
46Grok 4.5· LMArena data1,464.7
47Claude Opus 4.8· LMArena data1,464.6
48MiMo V2.5 Pro· LMArena data1,464.3
49Muse Spark 1.1· LMArena data1,463.6
50GLM 5.2Max· LMArena data1,462.6

Showing 1–50 of 194 results.

Data credits: LMArena, CC BY 4.0 (licence)