llm.ing

Design Arena — agenticgamedev

mirrorrun by Design Arenaunit: elo26 modelssource ↗

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

websitecodecategoriesgamedev3duicomponentdatavizsvgasciiartmobileappswebapps+19 more
#ModelScoreObserved
1Gemini 4 Argon
1266
Nov 17, 2026
2Claude Opus 4.7
1247
Sep 5, 2026
3Claude Fable 5
1231
Nov 17, 2026
4Claude Opus 5
1203
Nov 18, 2026
5Kimi K3
1202
Nov 18, 2026
6Claude Opus 4.8
1182
Nov 17, 2026
6Grok 4.5
1182
Nov 18, 2026
8Gemini 3.7 Flash
1181
Nov 18, 2026
9Claude Sonnet 5
1170
Nov 18, 2026
10GLM-5.2
1156
Nov 17, 2026
10Grok 4.6
1156
Nov 18, 2026
12Claude Opus 4.6
1144
Nov 18, 2026
13Muse Spark 1.1
1130
Nov 17, 2026
14Gemini 3.5 Flash
1127
Nov 18, 2026
15Muse Spark 1.2
1126
Nov 18, 2026
16GPT-5.5
1125
Nov 18, 2026
17Gemini 3.6 Flash
1124
Nov 18, 2026
18Claude Sonnet 4.6
1118
Nov 18, 2026
19MiniMax-M3
1113
Nov 18, 2026
20GLM-5.1
1109
Nov 18, 2026
21Qwen3.7 Max
1105
Nov 18, 2026
22Kimi K2.6
1088
Nov 18, 2026
23Kimi K2.7 Code
1076
Nov 17, 2026
24Gemini 3.1 Pro Preview
1061
Nov 18, 2026
25GLM-5V-Turbo
1049
Nov 18, 2026
26Grok 4.3
954
Nov 17, 2026

Dot = elo on a 938–1282 scale (position, not length). Scores link to model pages; every number keeps its source on the model page.