Skip to content

Benchmark

Arena ELO - Industry Entertainment and Sports and Media

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

Models measured
364
Scores recorded
421
Sources
1
IQ weight
1.17

Score distribution

1,279.8
Mean
1,314.7
Median
149.5
Std dev
1,058.5
10th percentile
1,441.1
90th percentile
784.8
Min
784.81,526.7
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.17 for chat models, measured over 364 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 data1,526.7
2Gemini 4 ArgonHigh· LMArena data1,522.1
3Claude Opus 5.5High· LMArena data1,506.5
4Gemini 4 HydrogenHigh· LMArena data1,499.5
5Claude Fable 5.1Max· LMArena data1,498.8
6Gemini 4 NeonHigh· LMArena data1,495.4
7Claude Opus 4.6High· LMArena data1,491.9
8Claude Fable 5.2Max· LMArena data1,486.2
9Claude Fable 5High· LMArena data1,483.9
10Gemini 3.8 FlashHigh· LMArena data1,482
11Muse WonderMax· LMArena data1,480.8
12Claude Opus 5Max· LMArena data1,480.4
13Claude Opus 4.6· LMArena data1,478.1
14Claude Opus 5High· LMArena data1,476.9
15Claude Opus 4.7High· LMArena data1,474.7
16Gemini 3.7 FlashHigh· LMArena data1,474.3
17Qwen3.8 Max· LMArena data1,471.5
18Claude Opus 4.7· LMArena data1,466.4
19MiMo V2.6 Pro· LMArena data1,465.7
20Muse Spark 1.3Max· LMArena data1,465.1
21Gemini 3.1 Pro Preview· LMArena data1,464.9
22Gemini 3 Pro Preview· LMArena data1,464.7
23Gemini 3.5 FlashHigh· LMArena data1,459.7
24Gemini 3.5 FlashMedium· LMArena data1,458.9
25Qwen3.5-Max-Preview· LMArena data1,456.8
26Gemini 3.6 FlashHigh· LMArena data1,456.6
27Kimi K3.1 ProMax· LMArena data1,454.9
28Claude Sonnet 5.5xHigh· LMArena data1,453.8
29Kimi K3Max· LMArena data1,451
30Muse Spark· LMArena data1,449.9
31Qwen3.7 Max Preview· LMArena data1,449.5
32GLM 5.3Max· LMArena data1,449.2
33Gemini 3 Flash· LMArena data1,448.9
34GPT-5.6 SolxHigh· LMArena data1,448.2
35Muse Spark 1.2xHigh· LMArena data1,448.1
36GLM 5.2Max· LMArena data1,447.9
37GPT 5.5High· LMArena data1,447.5
38GPT 5.5· LMArena data1,446.9
39GLM 5.1· LMArena data1,442.8
40DeepSeek V4.1 Flash 0910Max· LMArena data1,442.6
41GLM 5.3 Flash· LMArena data1,441.8
42Gemini 2.5 Pro· LMArena data1,441.4
43Claude Opus 4.8High· LMArena data1,441.1
44Muse Spark 1.1· LMArena data1,437.6
45GPT 5.4High· LMArena data1,437.5
46ERNIE 5.1· LMArena data1,437.1
47MiMo V2.5 Pro· LMArena data1,437
48Claude Opus 4.8· LMArena data1,434.4
49Claude Opus 4.520251101· LMArena data1,434
50Grok 4.5· LMArena data1,433.7

Showing 1–50 of 421 results.

Data credits: LMArena, CC BY 4.0 (licence)