Skip to content

Benchmark

Arena ELO Vision - Creative Writing Vision

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

Models measured
93
Scores recorded
106
Sources
1
IQ weight
1.74

Score distribution

1,246.2
Mean
1,254
Median
53.6
Std dev
1,181.7
10th percentile
1,318.8
90th percentile
1,100
Min
1,1001,342.1
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.74 for chat models, measured over 93 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 3 Pro Preview· LMArena data1,342.1
2Claude Fable 5High· LMArena data1,342
3Claude Opus 4.6High· LMArena data1,333.4
4Claude Opus 4.7High· LMArena data1,328.3
5Claude Fable 5.1Max· LMArena data1,327.4
6Gemini 3.1 Pro Preview· LMArena data1,324.4
7Claude Opus 5.5High· LMArena data1,323.9
8Muse Spark· LMArena data1,323.9
9Claude Opus 4.7· LMArena data1,323.6
10Claude Opus 4.6· LMArena data1,321
11Gemini 3.7 FlashHigh· LMArena data1,318.8
12Gemini 3 Flash· LMArena data1,314.8
13Gemini 3.8 FlashHigh· LMArena data1,313.8
14Claude Opus 5High· LMArena data1,307.7
15GPT 5.5High· LMArena data1,304.7
16Muse Spark 1.3Max· LMArena data1,303.8
17Qwen3.8 Max· LMArena data1,300.6
18Gemini 3.5 FlashHigh· LMArena data1,300.2
19Claude Opus 4.8High· LMArena data1,300
20Gemini 3.5 FlashMedium· LMArena data1,299
21GPT-6.1 SolMax· LMArena data1,297.4
22Gemini 3 FlashMinimal· LMArena data1,296
23Claude Opus 4.8· LMArena data1,292.8
24Grok 4.5· LMArena data1,291.3
25GPT 5.5· LMArena data1,289.5
26Gemma 4 31B IT· LMArena data1,286.6
27GPT 5.4High· LMArena data1,286.2
28GPT-5.6 SolxHigh· LMArena data1,285.6
29Muse Spark 1.3xHigh· LMArena data1,284.4
30DeepSeek V4.1 Flash 0910Max· LMArena data1,283.2
31Grok 4.20 Multi-Agent0309, beta· LMArena data1,281.7
32Grok 4.20Reasoning, 0309, beta· LMArena data1,276.6
33Claude Sonnet 5.5xHigh· LMArena data1,274.8
34Kimi K2.6· LMArena data1,273.6
35Claude Sonnet 4.6· LMArena data1,273.2
36GPT 5.4· LMArena data1,273.1
37GLM 5.3 Flash· LMArena data1,272.4
38Gemini 2.5 Flash Preview09-2025· LMArena data1,272.1
39Kimi K2.5Thinking· LMArena data1,270.7
40Gemini 3.6 FlashHigh· LMArena data1,269.9
41GPT 5.1· LMArena data1,266.9
42Gemini 2.5 Pro· LMArena data1,265.1
43GPT-6 AstraMax· LMArena data1,264.7
44Qwen3.7 Plus· LMArena data1,264.3
45Grok 4.6High· LMArena data1,263.3
46Qwen3.5 397B A17B· LMArena data1,262.3
47Doubao Seed 2.0 Pro· LMArena data1,258.7
48Gemma 4 26B A4B· LMArena data1,257.5
49Muse Spark 1.1· LMArena data1,257
50GPT-5.6 LunaxHigh· LMArena data1,256.2

Showing 1–50 of 106 results.

Data credits: LMArena, CC BY 4.0 (licence)