Skip to content

Benchmark

Design Arena Elo (models/ascii)

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

Models measured
76
Scores recorded
84
Sources
1
IQ weight
2.49

Score distribution

1,168.1
Mean
1,152
Median
97.1
Std dev
1,048
10th percentile
1,302
90th percentile
946
Min
9461,450
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.49 for chat models, measured over 76 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51GPT-5.6 Luna· Design Arena Leaderboards data1,135
52Qwen3.7 Plus· Design Arena Leaderboards data1,135
53DeepSeek V4 Pro· Design Arena Leaderboards data1,134
54Step 3.7 Flash· Design Arena Leaderboards data1,134
55GLM 5· Design Arena Leaderboards data1,133
56MiMo V2.5· Design Arena Leaderboards data1,128
57Claude Haiku 4.5· Design Arena Leaderboards data1,126
58GPT 5· Design Arena Leaderboards data1,126
59Qwen3 Max· Design Arena Leaderboards data1,113
60Grok 4.3· Design Arena Leaderboards data1,111
61GPT-5 mini· Design Arena Leaderboards data1,110
62GLM 5V Turbo· Design Arena Leaderboards data1,107
63GPT 5.1none· Design Arena Leaderboards data1,104
64GPT 5.1low· Design Arena Leaderboards data1,101
65Qwen3.6 Plus· Design Arena Leaderboards data1,098
66DeepSeek V4 Flash· Design Arena Leaderboards data1,095
67Grok 4 Fastreasoning· Design Arena Leaderboards data1,094
68Grok 4.1 Fastnon-reasoning· Design Arena Leaderboards data1,093
69GPT-5.1 Codex Mini· Design Arena Leaderboards data1,090
70DeepSeek V4 Flash 0731· Design Arena Leaderboards data1,083
71Inkling FP4· Design Arena Leaderboards data1,082
72Nemotron 3 Ultra· Design Arena Leaderboards data1,082
73Qwen3.5 Plus· Design Arena Leaderboards data1,079
74DeepSeek V3.2· Design Arena Leaderboards data1,068
75Mistral Large 3· Design Arena Leaderboards data1,055
76Mistral Medium 3.5· Design Arena Leaderboards data1,048
77GPT 5.1high· Design Arena Leaderboards data1,047
78Grok 4 Fastnon-reasoning· Design Arena Leaderboards data1,043
79Grok 4.1 Fastreasoning· Design Arena Leaderboards data1,040
80Trinity Large Thinking· Design Arena Leaderboards data1,027
81MiMo V2 Flash· Design Arena Leaderboards data1,003
82Mistral Medium 3.1· Design Arena Leaderboards data983
83Mercury 2· Design Arena Leaderboards data971
84INTELLECT-3· Design Arena Leaderboards data946

Showing 51–84 of 84 results.