Skip to content

Benchmark

Design Arena Elo (models/3d)

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

Models measured
135
Scores recorded
147
Sources
1
IQ weight
1.86

Score distribution

1,135.7
Mean
1,115
Median
150.4
Std dev
965
10th percentile
1,341
90th percentile
804
Min
8041,492
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.86 for chat models, measured over 135 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51GLM 5· Design Arena Leaderboards data1,196
52Claude Opus 4.8· Design Arena Leaderboards data1,189
53MiMo V2.5· Design Arena Leaderboards data1,185
54Claude Opus 4.5· Design Arena Leaderboards data1,181
55GPT-6 Luna· Design Arena Leaderboards data1,181
56GLM 5V Turbo· Design Arena Leaderboards data1,176
57MiniMax M3· Design Arena Leaderboards data1,176
58DeepSeek V4 Flash 0731· Design Arena Leaderboards data1,175
59Kimi K2.5· Design Arena Leaderboards data1,172
60Qwen3.6 Plus· Design Arena Leaderboards data1,170
61GPT 5.5· Design Arena Leaderboards data1,162
62Grok 4.20· Design Arena Leaderboards data1,160
63GLM 4.7· Design Arena Leaderboards data1,158
64DeepSeek V4 Flash· Design Arena Leaderboards data1,156
65Gemini 3 Flash· Design Arena Leaderboards data1,151
66Hy3 (Deep Infra)· Design Arena Leaderboards data1,141
67GLM 4.5· Design Arena Leaderboards data1,137
68MiniMax M2.1· Design Arena Leaderboards data1,132
69Qwen3.5 397B A17B· Design Arena Leaderboards data1,128
70Inkling FP4· Design Arena Leaderboards data1,127
71Claude Sonnet 4.5· Design Arena Leaderboards data1,124
72Claude Opus 4.1· Design Arena Leaderboards data1,120
73Claude Sonnet 4.5thinking· Design Arena Leaderboards data1,120
74DeepSeek V3.2 Exp· Design Arena Leaderboards data1,115
75DeepSeek V3.1thinking· Design Arena Leaderboards data1,108
76Claude Opus 4.1thinking, 20250805· Design Arena Leaderboards data1,107
77Claude Opus 4· Design Arena Leaderboards data1,106
78Claude Sonnet 4· Design Arena Leaderboards data1,106
79Nemotron 3.5 Super VL· Design Arena Leaderboards data1,103
80DeepSeek V3.2· Design Arena Leaderboards data1,100
81GPT-5.6 Terra· Design Arena Leaderboards data1,097
82GLM 4.6· Design Arena Leaderboards data1,096
83Nemotron 3 Ultra· Design Arena Leaderboards data1,096
84Step 3.7 Flash· Design Arena Leaderboards data1,096
85Claude Sonnet 4thinking· Design Arena Leaderboards data1,093
86Grok 4.3· Design Arena Leaderboards data1,091
87GLM 4.5 Air· Design Arena Leaderboards data1,090
88GLM 4.7 Flash· Design Arena Leaderboards data1,088
89GPT-5.6 Luna· Design Arena Leaderboards data1,086
90Qwen3.5 Plus· Design Arena Leaderboards data1,086
91MiMo V2 Flash· Design Arena Leaderboards data1,083
92Mistral Large 3· Design Arena Leaderboards data1,066
93Solar Mini 4· Design Arena Leaderboards data1,065
94Claude Sonnet 3.7· Design Arena Leaderboards data1,061
95Muse Glimmer 30B· Design Arena Leaderboards data1,059
96v0-1.5-md· Design Arena Leaderboards data1,058
97DeepSeek Chat· Design Arena Leaderboards data1,054
98INTELLECT-3· Design Arena Leaderboards data1,054
99Mistral Medium 3· Design Arena Leaderboards data1,051
100Gemini 2.5 Pro· Design Arena Leaderboards data1,048

Showing 51–100 of 147 results.