Skip to content

Benchmark

Arena ELO Vision

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

Models measured
147
Scores recorded
168
Sources
1
IQ weight
1.77

Score distribution

1,185.4
Mean
1,222.2
Median
114.5
Std dev
999
10th percentile
1,304.2
90th percentile
811.6
Min
811.61,325
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.77 for chat models, measured over 147 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Claude Fable 5High· LMArena data1,325
2Claude Fable 5.1Max· LMArena data1,322
3Claude Opus 5.5High· LMArena data1,321.8
4Claude Opus 4.7· LMArena data1,317.4
5Claude Opus 5High· LMArena data1,317.3
6Gemini 3.8 FlashHigh· LMArena data1,316.4
7Claude Opus 4.6High· LMArena data1,315.1
8Claude Opus 4.7High· LMArena data1,313.7
9Gemini 3.7 FlashHigh· LMArena data1,313.6
10Claude Fable 5.2Max· LMArena data1,313.3
11Claude Opus 4.6· LMArena data1,312.1
12Gemini 3.5 FlashHigh· LMArena data1,307.7
13Qwen3.8 Max· LMArena data1,306.7
14Muse Spark 1.3Max· LMArena data1,306.2
15Gemini 3.5 FlashMedium· LMArena data1,306.2
16Muse Spark· LMArena data1,305.6
17Gemini 3 Pro Preview· LMArena data1,304.6
18Muse Spark 1.2xHigh· LMArena data1,304.2
19GPT 5.4High· LMArena data1,300.9
20Muse Spark 1.3xHigh· LMArena data1,300.4
21Gemini 3.6 FlashHigh· LMArena data1,297
22Gemini 3.1 Pro Preview· LMArena data1,296.5
23GLM 5.3 Flash· LMArena data1,296.5
24GPT 5.5· LMArena data1,296
25GPT 5.5High· LMArena data1,295.2
26Claude Opus 4.8High· LMArena data1,294.3
27GPT 5.4· LMArena data1,291.3
28Muse Spark 1.1· LMArena data1,291.1
29Claude Opus 4.8· LMArena data1,288
30Grok 4.5· LMArena data1,287.8
31Claude Sonnet 5.5xHigh· LMArena data1,286
32Gemini 3 Flash· LMArena data1,284.8
33GPT-6.1 SolMax· LMArena data1,283.8
34Claude Sonnet 4.6· LMArena data1,282.5
35Kimi K2.6· LMArena data1,282.3
36GPT-5.6 SolxHigh· LMArena data1,282.2
37Gemma 4 31B IT· LMArena data1,279.3
38DeepSeek V4.1 Flash 0910Max· LMArena data1,278.3
39Qwen3.7 Plus· LMArena data1,277.3
40GPT-6 AstraMax· LMArena data1,276.2
41Doubao Seed 2.0 Pro· LMArena data1,274.1
42Claude Sonnet 5High· LMArena data1,271.5
43Qwen 3.8 27B· LMArena data1,271
44GPT-5.6 TerraxHigh· LMArena data1,269.8
45Kimi K2.5Thinking· LMArena data1,269.3
46Gemini 3.5 Flash Lite· LMArena data1,268.1
47GPT-5.2 Chat20260210· LMArena data1,267.6
48MiMo V2.6 Pro· LMArena data1,266
49Gemini 3 FlashMinimal· LMArena data1,265.6
50Grok 4.6High· LMArena data1,264.9

Showing 1–50 of 168 results.

Data credits: LMArena, CC BY 4.0 (licence)