Multimodal & Grounded benchmark
Design Arena Website leaderboard
Design Arena Website Elo. Every model the catalog carries a published Design Arena Website value for, ranked by that value.
CategoryMultimodal & Grounded
MeasureElo
TasksWebsite generation comparisons
DifficultyDesign and website generation
Published byOpenRouter Grok 4.3 benchmarks
A display-only Design Arena website-generation Elo score surfaced on OpenRouter model benchmark pages.
Design Arena Website ranking
88 models with a published Design Arena Website value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Elo |
|---|---|---|---|
| 1 |  Muse Spark 1.3 | Meta | 1364 |
| 2 |  Kimi K3 | Moonshot AI | 1353 |
| 3 |  Claude Fable 5.1 | Anthropic | 1322 |
| 4 | Muse Spark 1.2 | Meta | 1321 |
| 5 | Claude Opus 5 | Anthropic | 1319 |
| 6 |  GLM-5.3 | Z.AI | 1316 |
| 7 |  Gemini 3.7 Flash | 1315 | |
| 7 | Gemini 3.8 Flash | 1315 | |
| 9 | Gemini 3.6 Flash | 1313 | |
| 10 | Claude Fable 5 | Anthropic | 1308 |
| 11 | GLM-5.2 | Z.AI | 1305 |
| 12 |  Grok 4.6 | xAI | 1304 |
| 13 | Claude Opus 4.6 | Anthropic | 1303 |
| 13 | Claude Opus 4.6 (Adaptive) | Anthropic | 1303 |
| 15 | Claude Sonnet 4.6 | Anthropic | 1297 |
| 16 | Grok 4.5 | xAI | 1296 |
| 17 | GLM-5.1 | Z.AI | 1290 |
| 18 | Claude Sonnet 5 | Anthropic | 1289 |
| 19 | GLM-5.3-Flash | Z.AI | 1285 |
| 19 |  Qwen3.7 Max | Alibaba | 1285 |
| 21 |  MiMo-V2.5-Pro | Xiaomi | 1284 |
| 22 | Kimi K2.6 | Moonshot AI | 1282 |
| 22 | Qwen3.7 Plus | Alibaba | 1282 |
| 24 | Muse Spark 1.1 | Meta | 1281 |
| 25 | Kimi K2.7 Code | Moonshot AI | 1280 |
| 26 | MiMo-V2.5 | Xiaomi | 1279 |
| 27 | GLM-5-Turbo | Z.AI | 1278 |
| 28 | Gemini 3.5 Flash | 1275 | |
| 29 |  MiniMax M3 | MiniMax | 1271 |
| 30 |  GPT-5.5 | OpenAI | 1269 |
| 31 | Claude Opus 4.8 | Anthropic | 1267 |
| 32 | Gemini 3.1 Pro | 1265 | |
| 33 | Kimi K2.5 | Moonshot AI | 1261 |
| 33 | Kimi K2.5 (Reasoning) | Moonshot AI | 1261 |
| 35 | GLM-5 | Z.AI | 1260 |
| 35 | GLM-5 (Reasoning) | Z.AI | 1260 |
| 37 | Claude Opus 4.5 | Anthropic | 1259 |
| 37 | Claude Opus 4.5 Thinking | Anthropic | 1259 |
| 39 |  DeepSeek V4 Pro 0813 | DeepSeek | 1258 |
| 40 | MiniMax M2.7 | MiniMax | 1257 |
| 41 | Qwen3.6 Plus | Alibaba | 1253 |
| 42 | GLM-5V-Turbo | Z.AI | 1243 |
| 43 | Grok 4.20 | xAI | 1242 |
| 44 | GLM-4.7 | Z.AI | 1238 |
| 45 | GPT-5.4 | OpenAI | 1232 |
| 46 | TMInkling | Thinking Machines Lab | 1230 |
| 47 | DeepSeek V4 Flash 0731 | DeepSeek | 1220 |
| 48 | Gemini 3 Flash | 1208 | |
| 49 | GPT-5.2 | OpenAI | 1207 |
| 49 | Grok 4.3 | xAI | 1207 |
| 49 |  Step 3.7 Flash | StepFun | 1207 |
| 52 | Claude Sonnet 4.5 | Anthropic | 1202 |
| 52 | Claude Sonnet 4.5 Thinking | Anthropic | 1202 |
| 54 | GPT-5.1 | OpenAI | 1199 |
| 55 | GPT-5 (high) | OpenAI | 1197 |
| 55 | GPT-5 (medium) | OpenAI | 1197 |
| 57 |  Hy3 | Tencent | 1194 |
| 58 | Claude 4.1 Opus | Anthropic | 1189 |
| 59 |  Solar Pro 4 | Upstage | 1188 |
| 60 | DeepSeek V3.2 | DeepSeek | 1187 |
| 60 | DeepSeek V3.2 (Thinking) | DeepSeek | 1187 |
| 62 | GLM-4.5 | Z.AI | 1182 |
| 63 | Gemini 2.5 Pro | 1179 | |
| 64 | GPT-5.3 Codex | OpenAI | 1175 |
| 65 | GPT-5.1-Codex | OpenAI | 1174 |
| 66 | GLM-4.5-Air | Z.AI | 1159 |
| 67 | Claude 4 Sonnet | Anthropic | 1158 |
| 68 |  Trinity-Large-Preview | Arcee AI | 1149 |
| 68 | Trinity-Large-Thinking | Arcee AI | 1149 |
| 70 |  Nemotron 3 Ultra | NVIDIA | 1146 |
| 71 | Claude Haiku 4.5 | Anthropic | 1135 |
| 71 | Claude Haiku 4.5 Thinking | Anthropic | 1135 |
| 71 | DeepSeek V3.1 | DeepSeek | 1135 |
| 71 | DeepSeek V3.1 (Reasoning) | DeepSeek | 1135 |
| 75 | DeepSeek V3 | DeepSeek | 1132 |
| 76 | Qwen3 Max | Alibaba | 1131 |
| 77 | Gemini 2.5 Flash | 1127 | |
| 78 |  Mistral Medium 3 | Mistral | 1091 |
| 79 | Kimi K2 | Moonshot AI | 1063 |
| 80 | GPT-4.1 | OpenAI | 1051 |
| 81 | o3 | OpenAI | 1048 |
| 82 | GPT-4.1 mini | OpenAI | 1010 |
| 83 | GPT-4.1 nano | OpenAI | 985 |
| 84 | GPT-OSS 120B | OpenAI | 980 |
| 85 | Llama 4 Maverick | Meta | 883 |
| 86 | GPT-OSS 20B | OpenAI | 865 |
| 87 | GPT-4o | OpenAI | 843 |
| 88 | Llama 4 Scout | Meta | 762 |
Evidence key: Observed
Rows are ordered by the value OpenRouter Grok 4.3 benchmarks published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.