llm.ing

Design Arena — godotgamedev

mirrorrun by Design Arenaunit: elo35 modelssource ↗

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

websitegamedevcodecategories3duicomponentdatavizsvgasciiartmobileappswebapps+19 more
#ModelScoreObserved
1Claude Fable 5.1
1302
Nov 20, 2026
2Claude Opus 4.7
1270
Jul 23, 2026
3Claude Fable 5
1269
Nov 19, 2026
4Claude Opus 5
1256
Nov 20, 2026
5Grok 4.6
1253
Nov 19, 2026
6Gemini 3.8 Flash
1246
Nov 20, 2026
7Gemini 3.7 Flash
1234
Nov 20, 2026
8Claude Sonnet 4.6
1226
Jul 23, 2026
9Kimi K2.5
1219
Jul 23, 2026
10Muse Spark 1.3
1216
Nov 20, 2026
11Claude Opus 4.8
1208
Nov 20, 2026
12Grok 4.5
1204
Nov 19, 2026
13Kimi K3
1199
Nov 19, 2026
14Claude Sonnet 5
1198
Nov 19, 2026
15GLM-4.6
1180
Jul 23, 2026
16Qwen3.7 Max
1164
Nov 20, 2026
17Gemini 3 Flash Preview
1161
Jul 23, 2026
18GPT-5.5
1150
Nov 20, 2026
19GLM-5
1144
Jul 23, 2026
20GPT-5.2 Codex
1142
Jul 23, 2026
20GPT-5.2
1142
Jul 23, 2026
20GLM-5.2
1142
Jul 23, 2026
23Gemini 3.1 Pro Preview
1136
Nov 20, 2026
24GPT-5.4
1135
Jul 23, 2026
25GPT-5.3 Codex
1124
Jul 23, 2026
26Kimi K2.7 Code
1123
Nov 20, 2026
27Kimi K2.6
1098
Nov 19, 2026
28Grok 4.20 (Non-Reasoning)
1089
Jul 23, 2026
29Muse Spark 1.1
1081
Nov 19, 2026
30Gemini 3.5 Flash
1076
Nov 20, 2026
31DeepSeek V4 Pro
1059
Jul 23, 2026
32GLM-4.7
1058
Jul 23, 2026
33GLM-5.1
1033
Nov 20, 2026
34GLM-5V-Turbo
935
Nov 20, 2026
35Grok 4.3
931
Nov 20, 2026

Dot = elo on a 912–1321 scale (position, not length). Scores link to model pages; every number keeps its source on the model page.