Skip to content

Benchmark

Arena ELO Vision Style-Controlled

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

Models measured
147
Scores recorded
168
Sources
1
IQ weight
1.56

Score distribution

1,187.6
Mean
1,215.3
Median
96.7
Std dev
1,025.4
10th percentile
1,287.8
90th percentile
881.3
Min
881.31,308.8
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.56 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
101Gemini 2.5 Flash-Lite PreviewThinking, 06-17· LMArena data1,189
102Claude Sonnet 420250514· LMArena data1,189
103Qwen3 VL 235Bthinking· LMArena data1,188.1
104GPT-5 miniHigh· LMArena data1,185.6
105Qwen-VL Max2025-08-13· LMArena data1,184.3
106Grok 40709· LMArena data1,184
107Gemini 1.5 Pro002· LMArena data1,179.9
108Claude Sonnet 3.720250219· LMArena data1,175.6
109Gemini 2.5 Flash-Lite PreviewNon-Thinking, 09-2025· LMArena data1,173.8
110Gemini 2.0 Flash001· LMArena data1,172.4
111GPT 4o2024-05-13· LMArena data1,162.1
112Claude 3.5 Sonnet20241022· LMArena data1,161.4
113GLM 4.6v· LMArena data1,160.6
114Gemma 3 27B IT· LMArena data1,160.1
115Mistral Medium 3.12508· LMArena data1,159.5
116Mistral Medium 32505· LMArena data1,155.2
117Step-1o Turbo202506· LMArena data1,154.9
118Hunyuan Vision 1.5 Thinking· LMArena data1,153.8
119GLM 4.5V· LMArena data1,152.8
120Hunyuan Large Vision· LMArena data1,149.9
121GPT-5 nanoHigh· LMArena data1,147.9
122Claude 3.5 Sonnet20240620· LMArena data1,146
123Llama 4 Maverick· LMArena data1,145.6
124Step3· LMArena data1,143.8
125Gemini 1.5 Flash002· LMArena data1,142
126Mistral Small 3.22506· LMArena data1,140.1
127Gemini 2.0 Flash-Lite Preview02-05· LMArena data1,136.1
128Llama 4 Scout· LMArena data1,127.8
129Claude Haiku 3.520241022· LMArena data1,127.6
130Mistral Small 3.1 24B Instruct2503· LMArena data1,127
131Step-1o Vision 32KHigh Res· LMArena data1,123.9
132Qwen2.5 VL 72B Instruct· LMArena data1,121.3
133Gemini 1.5 Pro001· LMArena data1,119.8
134GPT 4o2024-08-06· LMArena data1,118.4
135Qwen 2.5 VL 32B Instruct· LMArena data1,118
136Molmo 2 8B· LMArena data1,112.6
137GPT-4 Turbo2024-04-09· LMArena data1,112.4
138GPT-4o mini2024-07-18· LMArena data1,097.1
139Pixtral Large2411· LMArena data1,095.5
140GPT-4.1 nano2025-04-14· LMArena data1,088.9
141Qwen2-VL-72B· LMArena data1,085.1
142Qwen-VL Max1119· LMArena data1,084.8
143Gemini 1.5 Flash-8B001· LMArena data1,071.7
144Step-1V 32K· LMArena data1,064.4
145Claude 3 Opus20240229· LMArena data1,062.3
146Gemini 1.5 Flash001· LMArena data1,059.6
147Molmo 72B0924· LMArena data1,045.6
148Hunyuan Standard Vision2024-12-31· LMArena data1,042.1
149Llama 3.2 90B Vision Instruct· LMArena data1,032.7
150Qwen2 VL 7b Instruct· LMArena data1,031.2

Showing 101–150 of 168 results.

Data credits: LMArena, CC BY 4.0 (licence)