llm.ing

Design Arena — mobileapps

mirrorrun by Design Arenaunit: elo30 modelssource ↗

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

websitecodecategoriesdatavizuicomponentgamedev3dsvgasciiartwebappsfullstack+15 more
#ModelScoreObserved
1Claude Opus 4.8
1281
Jul 23, 2026
2Claude Opus 4.6
1280
Jul 23, 2026
3Claude Sonnet 4.6
1272
Jul 23, 2026
4Claude Fable 5
1260
Jul 24, 2026
5Claude Opus 4.5 (latest)
1254
Jul 23, 2026
5MiniMax-M3
1254
Jul 24, 2026
7Gemini 3.5 Flash
1250
Jul 23, 2026
8Kimi K2.6
1248
Jul 23, 2026
9GLM-5.2
1238
Jul 24, 2026
10GLM-5.1
1226
Jul 24, 2026
11Kimi K2.7 Code
1220
Jul 23, 2026
12GPT-5.5
1219
Jul 24, 2026
13Claude Sonnet 4.5 (latest)
1215
Jul 23, 2026
14GLM-5
1214
Jul 23, 2026
15Qwen3.7 Max
1209
Jul 24, 2026
16GLM-5V-Turbo
1205
Jul 23, 2026
17GPT-5.1 Codex
1198
Jul 24, 2026
18GLM-4.7
1184
Jul 23, 2026
19Grok 4.20 (Non-Reasoning)
1182
Jul 23, 2026
20Kimi K2.5
1181
Jul 23, 2026
21Gemini 3 Flash Preview
1179
Jul 24, 2026
21GLM-4.6
1179
Jul 24, 2026
23Gemini 3.1 Pro Preview
1168
Jul 24, 2026
24GPT-5.2
1167
Jul 23, 2026
25GPT-5.2 Codex
1165
Jul 23, 2026
26GPT-5.4
1147
Jul 23, 2026
27GPT-5-Codex
1132
Jul 23, 2026
28Grok 4.3
1130
Jul 23, 2026
29GPT-5.3 Codex
1128
Jul 24, 2026
30GPT-5.1
1123
Jul 23, 2026

Dot = elo on a 1115–1289 scale (position, not length). Scores link to model pages; every number keeps its source on the model page.