llm.ing

Design Arena — codecategories

mirrorrun by Design Arenaunit: elo81 modelssource ↗

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

websitecodecategoriesdatavizuicomponentgamedev3dsvgasciiartfullstackmobileapps+15 more
#ModelScoreObserved
1Kimi K3
1384
Sep 21, 2026
2Muse Spark 1.3
1369
Sep 21, 2026
3Claude Fable 5.1
1344
Sep 21, 2026
4Claude Opus 5
1336
Sep 21, 2026
5GLM-5.3
1326
Sep 21, 2026
6Claude Fable 5
1324
Sep 21, 2026
6Muse Spark 1.2
1324
Sep 21, 2026
8Gemini 3.7 Flash
1322
Sep 21, 2026
8Gemini 3.8 Flash
1322
Sep 21, 2026
10Qwen3.8 Max
1311
Sep 4, 2026
11Claude Opus 4.7
1307
Sep 5, 2026
11Grok 4.6
1307
Sep 21, 2026
13GLM-5.2
1306
Sep 21, 2026
14Gemini 3.6 Flash
1303
Sep 21, 2026
15Claude Opus 4.6
1302
Sep 21, 2026
16GLM-5.3-Flash
1296
Sep 21, 2026
17Claude Sonnet 5
1291
Sep 21, 2026
17Grok 4.5
1291
Sep 21, 2026
17Muse Spark 1.1
1291
Sep 21, 2026
20Claude Sonnet 4.6
1290
Sep 21, 2026
21Qwen3.7 Max
1289
Sep 21, 2026
22Kimi K2.6
1286
Sep 21, 2026
23Qwen3.7 Plus
1282
Sep 21, 2026
24GLM-5.1
1281
Sep 21, 2026
25Gemini 3.5 Flash
1277
Sep 21, 2026
26GPT-5.5
1273
Sep 21, 2026
27Kimi K2.7 Code
1269
Sep 21, 2026
28Claude Opus 4.8
1265
Sep 21, 2026
29MiniMax-M3
1263
Sep 21, 2026
30GLM-5
1260
Sep 21, 2026
31Gemini 3.1 Pro Preview
1257
Sep 21, 2026
32Claude Opus 4.5 (latest)
1256
Sep 21, 2026
33DeepSeek V4 Pro
1254
Sep 21, 2026
34Kimi K2.5
1250
Sep 21, 2026
35Qwen3.6 Plus
1249
Sep 21, 2026
36MiniMax-M2.7
1247
Sep 21, 2026
37DeepSeek V4 Flash
1244
Sep 21, 2026
38GLM-5V-Turbo
1242
Sep 21, 2026
39Grok 4.20 (Non-Reasoning)
1234
Sep 21, 2026
40GLM-4.7
1230
Sep 21, 2026
41GPT-5.4
1225
Sep 21, 2026
42MiniMax-M2.5
1221
Sep 21, 2026
43MiniMax-M2.1
1204
Sep 21, 2026
44Gemini 3 Flash Preview
1203
Sep 21, 2026
45Grok 4.3
1200
Sep 21, 2026
46Claude Sonnet 4.5 (latest)
1196
Sep 21, 2026
47Step 3.7 Flash
1194
Sep 21, 2026
48Qwen3.5 397B-A17B
1193
Sep 21, 2026
49GLM-4.7-Flash
1191
Sep 21, 2026
50Claude Opus 4.1 (latest)
1184
Sep 21, 2026
50GPT-5.1
1184
Sep 21, 2026
52GPT-5.2
1183
Sep 21, 2026
53GPT-5
1181
Sep 21, 2026
54Qwen3.5 Plus
1180
Sep 21, 2026
55GLM-4.5
1179
Sep 21, 2026
55GLM-4.6
1179
Sep 21, 2026
57DeepSeek V3.2
1178
Sep 21, 2026
58GPT-5.1 Codex
1166
Sep 21, 2026
58Gemini 2.5 Pro
1166
Sep 21, 2026
60GPT-5.3 Codex
1158
Sep 21, 2026
61Nemotron 3 Ultra 550B A55B
1156
Sep 21, 2026
62DeepSeek R1
1152
Sep 21, 2026
63GLM-4.5-Air
1151
Sep 21, 2026
64MiniMax-M2
1149
Sep 21, 2026
65Claude Haiku 4.5 (latest)
1128
Sep 21, 2026
65GPT-5 Mini
1128
Sep 21, 2026
67Qwen3 Max
1123
Sep 21, 2026
68DeepSeek Chat
1122
Sep 21, 2026
69Gemini 2.5 Flash
1118
Sep 21, 2026
70Kimi K2 Thinking
1111
Sep 21, 2026
71GPT-5.1 Codex mini
1107
Sep 21, 2026
72GPT-5 Nano
1098
Sep 21, 2026
73Gemini 3.1 Flash Lite Preview
1086
Sep 21, 2026
74Qwen3 235B-A22B
1052
Sep 21, 2026
75GPT-4.1
1039
Sep 21, 2026
76o3
1033
Sep 21, 2026
77Mercury 2
1027
Sep 21, 2026
78GPT-4.1 mini
1007
Sep 21, 2026
79o4-mini
989
Sep 21, 2026
80GPT-4.1 nano
976
Sep 21, 2026
81GPT-4o
874
Sep 21, 2026

Dot = elo on a 848–1410 scale (position, not length). Scores link to model pages; every number keeps its source on the model page.