llm.ing

Design Arena — dataviz

mirrorrun by Design Arenaunit: elo81 modelssource ↗

Caveat: Head-to-head human votes on generated designs/code, mirrored via OpenRouter.

websitecodecategoriesdatavizuicomponentgamedev3dsvgasciiartfullstackmobileapps+15 more
#ModelScoreObserved
1GLM-5.1
1366
Jul 23, 2026
2Muse Spark 1.2
1365
Sep 19, 2026
3Claude Fable 5.1
1364
Sep 20, 2026
4Kimi K3
1358
Sep 20, 2026
5Claude Opus 5
1352
Sep 21, 2026
6Muse Spark 1.3
1348
Sep 21, 2026
7Claude Fable 5
1323
Sep 21, 2026
7Gemini 3.7 Flash
1323
Sep 21, 2026
9Gemini 3.6 Flash
1310
Sep 20, 2026
10GLM-5.2
1309
Sep 20, 2026
11Claude Opus 4.7
1306
Sep 4, 2026
12Grok 4.6
1302
Sep 20, 2026
13Claude Opus 4.6
1296
Sep 21, 2026
14Claude Sonnet 4.6
1294
Sep 20, 2026
15Qwen3.7 Max
1292
Sep 20, 2026
16Qwen3.7 Plus
1291
Sep 21, 2026
17Muse Spark 1.1
1290
Sep 21, 2026
18Grok 4.5
1285
Sep 20, 2026
19GLM-5.3-Flash
1277
Sep 21, 2026
20Qwen3.8 Max
1276
Sep 4, 2026
21GPT-5.5
1275
Sep 21, 2026
22Gemini 3.8 Flash
1272
Sep 21, 2026
23GLM-5.3
1270
Sep 21, 2026
24Kimi K2.6
1265
Sep 18, 2026
25Claude Sonnet 5
1261
Sep 20, 2026
26Claude Opus 4.5 (latest)
1258
Sep 16, 2026
27Claude Opus 4.8
1254
Sep 21, 2026
27Gemini 3.1 Pro Preview
1254
Sep 20, 2026
29GPT-5.4
1250
Sep 19, 2026
29MiniMax-M2.7
1250
Sep 20, 2026
31MiniMax-M3
1248
Sep 21, 2026
32Gemini 3.5 Flash
1247
Sep 20, 2026
33Qwen3.6 Plus
1244
Sep 20, 2026
33GLM-5
1244
Sep 21, 2026
35GPT-5
1236
Sep 17, 2026
35Gemini 2.5 Pro
1236
Sep 17, 2026
35Kimi K2.7 Code
1236
Sep 21, 2026
38Kimi K2.5
1234
Sep 17, 2026
39Grok 4.20 (Non-Reasoning)
1233
Sep 17, 2026
40GPT-5.1 Codex
1228
Sep 17, 2026
41MiniMax-M2.1
1221
Sep 18, 2026
42GPT-5.1
1220
Sep 20, 2026
43DeepSeek V4 Pro
1218
Sep 18, 2026
44GPT-5.2
1216
Sep 16, 2026
45GLM-5V-Turbo
1215
Sep 17, 2026
46GLM-4.7
1211
Sep 20, 2026
47GLM-4.5-Air
1201
Sep 17, 2026
48o3
1200
Jul 23, 2026
49DeepSeek R1
1197
Sep 20, 2026
49Grok 4.3
1197
Sep 20, 2026
51DeepSeek V4 Flash
1196
Sep 20, 2026
52Qwen3.5 397B-A17B
1190
Sep 17, 2026
53MiniMax-M2.5
1187
Sep 16, 2026
53Step 3.7 Flash
1187
Sep 20, 2026
55Claude Sonnet 4.5 (latest)
1182
Sep 17, 2026
56Claude Opus 4.1 (latest)
1181
Sep 16, 2026
57GPT-5.3 Codex
1179
Sep 17, 2026
58GLM-4.6
1176
Sep 17, 2026
59DeepSeek V3.2
1174
Sep 20, 2026
59GLM-4.5
1174
Sep 16, 2026
61Nemotron 3 Ultra 550B A55B
1161
Sep 21, 2026
62MiniMax-M2
1156
Sep 21, 2026
63Qwen3.5 Plus
1151
Sep 16, 2026
64Gemini 2.5 Flash
1149
Sep 20, 2026
65GPT-5 Mini
1145
Sep 17, 2026
66Claude Haiku 4.5 (latest)
1140
Sep 21, 2026
67GLM-4.7-Flash
1138
Sep 17, 2026
68GPT-4.1
1117
Sep 17, 2026
69Qwen3 Max
1115
Sep 16, 2026
70GPT-5.1 Codex mini
1112
Sep 16, 2026
71DeepSeek Chat
1103
Sep 18, 2026
72Qwen3-Coder 30B-A3B Instruct
1098
Sep 20, 2026
73Qwen3 235B-A22B
1084
Sep 17, 2026
74GPT-5 Nano
1071
Sep 20, 2026
75Gemini 3.1 Flash Lite Preview
1060
Sep 16, 2026
76GPT-4.1 mini
1047
Sep 17, 2026
77Kimi K2 Thinking
1031
Sep 17, 2026
78Mercury 2
1015
Sep 21, 2026
79o4-mini
1005
Sep 16, 2026
80GPT-4.1 nano
905
Sep 16, 2026
81GPT-4o
870
Sep 17, 2026

Dot = elo on a 845–1391 scale (position, not length). Scores link to model pages; every number keeps its source on the model page.