llm.ing

Design Arena — htmlslides

mirrorrun by Design Arenaunit: elo18 modelssource ↗

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

websitecodecategoriesdatavizuicomponentgamedev3dsvgasciiartwebappsfullstack+15 more
#ModelScoreObserved
1Claude Fable 5
1260
Jul 24, 2026
2Claude Opus 4.7
1241
Jul 24, 2026
3Claude Opus 4.8
1235
Jul 23, 2026
4Grok 4.5
1234
Jul 24, 2026
5Claude Sonnet 5
1233
Jul 23, 2026
6Kimi K2.6
1229
Jul 24, 2026
7Muse Spark 1.1
1223
Jul 24, 2026
8Kimi K2.7 Code
1220
Jul 24, 2026
9Gemini 3.1 Pro Preview
1204
Jul 24, 2026
9GLM-5.2
1204
Jul 24, 2026
11GLM-5.1
1201
Jul 23, 2026
12Qwen3.7 Max
1191
Jul 24, 2026
13Grok 4.20 (Non-Reasoning)
1181
Jul 24, 2026
14GLM-5
1172
Jul 24, 2026
15Gemini 3.5 Flash
1170
Jul 24, 2026
16GLM-5V-Turbo
1136
Jul 24, 2026
17GPT-5.5
1086
Jul 24, 2026
18Grok 4.3
1047
Jul 23, 2026

Dot = elo on a 1036–1271 scale (position, not length). Scores link to model pages; every number keeps its source on the model page.