Arena Code WebDev — Reference-Based Design
Code
Arena Code WebDev rating for reference-based design tasks. Anonymous generated experiences are compared through blind pairwise user votes.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Opus 5 | 1742 |
| Kimi K3 | 1720 |
| Claude Fable 5 | 1657 |
| Opus 4.7 | 1583 |
| GLM-5.2 | 1580 |
| Opus 4.8 | 1577 |
| DeepSeek-V4-Flash | 1573 |
| Grok 4.5 | 1572 |
| Opus 4.6 | 1565 |
| Muse Spark 1.1 | 1558 |
| Sonnet 5 | 1550 |
| Gemini 3.6 Flash | 1547 |
| GLM-5.1 | 1538 |
| Sonnet 4.6 | 1535 |
| Hy3 | 1517 |
| Kimi K2.6 | 1512 |
| Gemini 3.5 Flash | 1500 |
| MiniMax M3 | 1489 |
| Kimi K2.7 Code | 1474 |
| MiMo-V2.5-Pro | 1472 |
| Opus 4.5 | 1472 |
| Qwen3.6 Plus (2026-04-02) | 1463 |
| Gemini 3.1 Pro Preview | 1454 |
| DeepSeek-V4-Pro | 1453 |
| Gemini 3 Pro Preview | 1451 |
| MiMo-V2.5 | 1449 |
| GLM-5 | 1448 |
| Gemini 3 Flash Preview | 1444 |
| MiMo-V2-Pro | 1443 |
| Kimi K2.5 | 1438 |
| GLM-4.7 | 1437 |
| GPT-5 | 1424 |
| GPT-5.3-Codex | 1416 |
| Opus 4.1 | 1415 |
| GPT-5.4 | 1413 |
| GPT-5.4 Mini | 1411 |
| Qwen3.5 397B A17B | 1399 |
| Inkling | 1399 |
| GPT-5.2 | 1396 |
| MiniMax M2.7 | 1394 |