Arena Image-to-WebDev
Code
Arena Image-to-WebDev ranks models that turn visual references into working web experiences. Anonymous generated results are compared through blind pairwise user votes.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Sonnet 5.5 | 1737 |
| GPT-6 Astra | 1726 |
| Claude Fable 5.1 | 1722 |
| GPT-6.1 Sol | 1702 |
| Claude Opus 5 | 1662 |
| GPT-6 Sol | 1660 |
| Muse Spark 1.3 | 1640 |
| Qwen3.8 Max (0902) | 1637 |
| Claude Fable 5 | 1623 |
| Qwen3.8-Max | 1614 |
| GPT-5.6 Sol | 1609 |
| DeepSeek V4.1 Flash | 1606 |
| Grok 4.6 | 1603 |
| Gemini 3.7 Flash | 1599 |
| MiMo-V2.6-Pro | 1598 |
| MiMo-V2.6-Flash | 1596 |
| Gemini 3.8 Flash | 1593 |
| Grok 4.7 | 1586 |
| Kimi K3 | 1579 |
| GLM-5.3 Flash | 1578 |
| Opus 4.7 | 1577 |
| Qwen3.8 27B | 1577 |
| GPT-6 Luna | 1571 |
| Grok 4.5 | 1562 |
| Gemini 3.6 Flash | 1545 |
| Opus 4.6 | 1543 |
| GPT-5.6 Terra | 1541 |
| Sonnet 5 | 1540 |
| Muse Spark 1.1 | 1540 |
| Sonnet 4.6 | 1529 |
| Kimi K2.6 | 1521 |
| GPT-5.6 Luna | 1517 |
| Kimi K2.7 Code | 1513 |
| Gemini 3.5 Flash | 1506 |
| Gemini 3.1 Pro Preview | 1481 |
| Qwen3.6 Plus (2026-04-02) | 1472 |
| MiniMax M3 | 1470 |
| Gemini 3 Flash Preview | 1450 |
| Gemini 3.5 Flash-Lite | 1448 |
| Gemini 3 Pro Preview | 1441 |