Benchmarks · coding
Arena WebDev · HTML
Human preference between agentic front-end web development outputs; Bradley-Terry rating. Only the HTML category.
Category
112 models · 122 results · data as of 2026-10-29 · arena.ai ↗
Leaderboard
GPT-6.1 Sol max
All results
122 rows| # | Model | Effort | Arena WebDev · HTML |
|---|---|---|---|
| 1 | GPT-6.1 Sol | max | 1796 ±28 |
| 2 | GPT-6 Astra | max | 1785 ±21 |
| 3 | Gemini 4 Carbon | high | 1780 ±48 |
| 4 | Opus 5.5 | max | 1753 ±53 |
| 5 | Sonnet 5.5 | xhigh | 1739 ±30 |
| 6 | GPT-6.1 Terra | max | 1719 ±37 |
| 7 | Kimi K3.1 | max | 1706 ±28 |
| 8 | DeepSeek V4.1 Flash | max | 1701 ±18 |
| 9 | Sonnet 5.5 | high | 1700 ±22 |
| 10 | Fable 5.1 | max | 1683 ±22 |
| 11 | GPT-6 Sol | max | 1680 ±22 |
| 12 | Opus 5 | max | 1677 ±14 |
| 13 | Gemini 4 Argon | high | 1667 ±37 |
| 14 | Qwen 3.8 Max | max | 1657 ±13 |
| 15 | Muse Spark 1.3 | max | 1654 ±17 |
| 16 | Fable 5 | high | 1651 ±14 |
| 17 | Opus 5 | high | 1649 ±11 |
| 18 | Kimi K3 | max | 1641 ±13 |
| 19 | MiMo V2.6 Flash | high | 1638 ±19 |
| 20 | Grok 4.7 | xhigh | 1629 ±20 |
| 21 | MiMo V2.6 Pro | high | 1624 ±20 |
| 22 | Hy4 | high | 1620 ±25 |
| 23 | Muse Spark 1.3 | xhigh | 1620 ±17 |
| 24 | GPT-5.6 Sol | max | 1617 ±13 |
| 25 | Gemini 3.8 Flash | high | 1612 ±16 |
| 26 | Grok 4.6 | high | 1597 ±15 |
| 27 | GLM 5.3 | max | 1597 ±17 |
| 28 | Qwen 3.8 Flash | max | 1595 ±16 |
| 29 | GPT-6 Luna | max | 1595 ±11 |
| 30 | Gemini 3.7 Flash | high | 1591 ±14 |