llm.ing

Design Arena — pptxslides

mirrorrun by Design Arenaunit: eloobserved Jul 23, 20269 modelssource ↗

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

websitecodecategoriesdatavizuicomponentgamedev3dsvgasciiartwebappsfullstack+15 more
#ModelScore
1Claude Opus 4.7
1333
2Claude Opus 4.8
1306
3Gemini 3.5 Flash
1244
4GLM-5.1
1241
5Kimi K2.6
1181
6GLM-5V-Turbo
1164
7GPT-5.5
1157
8Gemini 3.1 Pro Preview
1110
9Grok 4.3
1071

Dot = elo on a 1057–1347 scale (position, not length). Scores link to model pages; every number keeps its source on the model page.