Skip to content

Benchmark

Arena ELO Vision - Diagram

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

Models measured
108
Scores recorded
124
Sources
1
IQ weight
1.87

Score distribution

1,253.6
Mean
1,264.9
Median
64
Std dev
1,152.7
10th percentile
1,322.5
90th percentile
1,074
Min
1,0741,354.2
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.87 for chat models, measured over 108 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 data1,354.2
2Claude Opus 5.5High· LMArena data1,351.9
3Gemini 3.8 FlashHigh· LMArena data1,347.4
4Claude Fable 5.1Max· LMArena data1,345.3
5Claude Opus 5High· LMArena data1,344.6
6Claude Opus 4.6· LMArena data1,339.4
7Claude Opus 4.7High· LMArena data1,336.7
8Muse Spark 1.3Max· LMArena data1,336.1
9Claude Opus 4.7· LMArena data1,334.6
10Claude Opus 4.6High· LMArena data1,331.6
11Gemini 3.7 FlashHigh· LMArena data1,329.3
12GPT 5.4High· LMArena data1,325.5
13GPT 5.5· LMArena data1,322.5
14Gemini 3.5 FlashMedium· LMArena data1,322.1
15Gemini 3.5 FlashHigh· LMArena data1,321.9
16Claude Fable 5.2Max· LMArena data1,321.6
17Qwen3.8 Max· LMArena data1,320.4
18GLM 5.3 Flash· LMArena data1,320.3
19Claude Sonnet 5.5xHigh· LMArena data1,317.8
20GPT 5.5High· LMArena data1,317.6
21Gemini 3.6 FlashHigh· LMArena data1,316.8
22GPT-6.1 SolMax· LMArena data1,316
23GPT 5.4· LMArena data1,315.6
24Muse Spark 1.3xHigh· LMArena data1,314.1
25Muse Spark 1.2xHigh· LMArena data1,314.1
26Claude Opus 4.8High· LMArena data1,313.5
27Gemini 3.1 Pro Preview· LMArena data1,311.4
28Muse Spark· LMArena data1,309
29GPT-5.6 SolxHigh· LMArena data1,306.3
30Muse Spark 1.1· LMArena data1,305.4
31Claude Sonnet 4.6· LMArena data1,305.2
32Grok 4.5· LMArena data1,304.6
33GPT-6 AstraMax· LMArena data1,302.6
34Claude Opus 4.8· LMArena data1,302.2
35Gemini 3 Pro Preview· LMArena data1,301.7
36Step 5 Preview· LMArena data1,300.7
37Kimi K2.6· LMArena data1,298.6
38DeepSeek V4.1 Flash 0910Max· LMArena data1,296.1
39Claude Sonnet 5High· LMArena data1,293.1
40GPT-5.6 TerraxHigh· LMArena data1,291.8
41Gemini 3 Flash· LMArena data1,289.5
42Doubao Seed 2.0 Pro· LMArena data1,289.3
43Qwen3.7 Plus· LMArena data1,289
44Grok 4.6High· LMArena data1,286.8
45Gemma 4 31B IT· LMArena data1,286.8
46MiMo V2.6 Pro· LMArena data1,285.6
47GPT-5.2 Chat20260210· LMArena data1,284.7
48Kimi K2.5Thinking· LMArena data1,282.6
49Qwen 3.8 27B· LMArena data1,280.4
50Gemini 3.5 Flash Lite· LMArena data1,279

Showing 1–50 of 124 results.

Data credits: LMArena, CC BY 4.0 (licence)