Skip to content

Benchmark

Arena ELO - Polish

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

Models measured
207
Scores recorded
242
Sources
1
IQ weight
2.07

Score distribution

1,386.8
Mean
1,405.7
Median
84.1
Std dev
1,270
10th percentile
1,481.7
90th percentile
1,072
Min
1,0721,546.7
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.07 for chat models, measured over 207 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 data1,546.7
2Claude Fable 5.1Max· LMArena data1,536.2
3Muse WonderMax· LMArena data1,535.2
4Claude Opus 5Max· LMArena data1,522.6
5Claude Opus 5High· LMArena data1,516.8
6Gemini 3.5 FlashHigh· LMArena data1,515.9
7Claude Opus 5.5High· LMArena data1,514.7
8Gemini 3.8 FlashHigh· LMArena data1,510
9Muse Spark 1.2xHigh· LMArena data1,508.5
10Gemini 3.7 FlashHigh· LMArena data1,506.4
11Gemini 3 Pro Preview· LMArena data1,503
12Muse Spark 1.1· LMArena data1,501.6
13Gemini 3.5 FlashMedium· LMArena data1,499.1
14Gemini 3.1 Pro Preview· LMArena data1,497.9
15Muse Spark 1.3Max· LMArena data1,497.5
16Claude Opus 4.6High· LMArena data1,495.7
17Claude Sonnet 5.5xHigh· LMArena data1,494.8
18Gemini 3 Flash· LMArena data1,493.7
19Claude Opus 4.7· LMArena data1,491.4
20Claude Fable 5High· LMArena data1,488.5
21Claude Opus 4.6· LMArena data1,487.6
22Doubao Seed 2.0 Pro· LMArena data1,487.5
23ERNIE-5.0-Preview1203· LMArena data1,485.2
24ERNIE-5.0-Preview1022· LMArena data1,482.8
25Qwen3.8 Max· LMArena data1,481.7
26Claude Opus 4.7High· LMArena data1,479.5
27Gemini 2.5 Pro· LMArena data1,479.1
28Gemini 3.6 FlashHigh· LMArena data1,478.9
29Muse Spark· LMArena data1,474.2
30Kimi K3Max· LMArena data1,473.2
31GPT 5.4High· LMArena data1,472.3
32Gemini 3 FlashMinimal· LMArena data1,469.3
33ERNIE 5.1· LMArena data1,468
34ERNIE 5.00110· LMArena data1,467.6
35GPT 5.5High· LMArena data1,464.2
36GLM 5.3Max· LMArena data1,463.2
37GPT-5.6 SolxHigh· LMArena data1,463
38MiMo V2.5 Pro· LMArena data1,461
39Claude Opus 4.8High· LMArena data1,459.7
40GLM 5.2Max· LMArena data1,459.4
41GPT 5.5· LMArena data1,459.2
42Claude Opus 4.8· LMArena data1,458.8
43DeepSeek V4 Pro· LMArena data1,458.5
44Qwen3.5-Max-Preview· LMArena data1,456.6
45Grok 4.1Thinking· LMArena data1,455.1
46Grok 4.20 Multi-Agent0309, beta· LMArena data1,454.6
47GPT 5.4· LMArena data1,454.6
48DeepSeek V4.1 Flash 0910Max· LMArena data1,453.1
49DeepSeek-V4-Pro PreviewHigh· LMArena data1,452.3
50GLM 5.1· LMArena data1,452.3

Showing 1–50 of 242 results.

Data credits: LMArena, CC BY 4.0 (licence)