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
51Qwen3.5 397B A17B· LMArena data1,277
52Gemini 3 FlashMinimal· LMArena data1,276
53GPT-5.5 Instant· LMArena data1,273.5
54Gemini 2.5 Pro· LMArena data1,273.1
55Mistral Large 4· LMArena data1,272.6
56GPT-5.6 LunaxHigh· LMArena data1,272.1
57MiMo V2.5· LMArena data1,269.5
58MiMo V2.6 Flash· LMArena data1,269.2
59Grok 4.20Reasoning, 0309, beta· LMArena data1,267.1
60GPT 5.2High· LMArena data1,265.3
61Qwen-VL Max2025-08-13· LMArena data1,264.9
62MiniMax M3· LMArena data1,264.9
63GPT-5.4 miniHigh· LMArena data1,264.9
64GPT 5.1High· LMArena data1,264.2
65Gemma 4 26B A4B· LMArena data1,264.1
66GPT-6 SolMax· LMArena data1,263.5
67Grok 4.20 Multi-Agent0309, beta· LMArena data1,261.2
68GPT-6.1 TerraMax· LMArena data1,258.9
69Kimi K2.5Instant· LMArena data1,257.8
70Grok 4.7xHigh· LMArena data1,257.5
71MiMo V2 Omni· LMArena data1,256.3
72Mistral Medium 3.5· LMArena data1,255.4
73Qwen3 VL 235B· LMArena data1,255.2
74ChatGPT-4o Latest20250326· LMArena data1,254
75Inkling Small· LMArena data1,253.5
76GPT-5 Chat· LMArena data1,251.1
77Qwen3.5 27B· LMArena data1,249.8
78Qwen3.5 122B-A10B· LMArena data1,249.1
79Gemini 2.5 Flash Preview09-2025· LMArena data1,247.6
80GPT 5.2· LMArena data1,245.5
81GPT 5.1· LMArena data1,245.1
82Gemini 2.5 Flash· LMArena data1,245.1
83Gemini 3.1 Flash Lite· LMArena data1,242.1
84Grok 4.3· LMArena data1,241.1
85Mistral Large 3· LMArena data1,238.4
86GPT-6 LunaMax· LMArena data1,237.8
87ERNIE-5.0-Preview1220· LMArena data1,235.5
88NVIDIA Nemotron 3.5 Super VL 120BNVFP4· LMArena data1,226.9
89GPT 5High· LMArena data1,225.9
90OpenAI o32025-04-16· LMArena data1,219.4
91GPT 4.12025-04-14· LMArena data1,214.3
92Qwen3 VL 235Bthinking· LMArena data1,211.4
93Grok 4.1 FastReasoning· LMArena data1,211.4
94GPT 5.4 NanoHigh· LMArena data1,210
95Grok 40709· LMArena data1,203.5
96GPT-5 miniHigh· LMArena data1,202.3
97Claude Sonnet 4Thinking 32K, 20250514· LMArena data1,198.8
98o4 Mini2025-04-16· LMArena data1,196.7
99Gemini 2.5 Flash-Lite PreviewNon-Thinking, 09-2025· LMArena data1,195.2
100Hunyuan Vision 1.5 Thinking· LMArena data1,192.2

Showing 51–100 of 124 results.

Data credits: LMArena, CC BY 4.0 (licence)