Web-Bench: leaderboard
Web development benchmark testing code generation for realistic web applications with Pass@1 evaluation across multiple frameworks.
Metric: Pass@1 (%). Source: huggingface.co. Saturation forecast: Around 2029. 48 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4 (20250514) | 28 | #154 |
| 2 | Claude Opus 4 (20250514) (Thinking) | 25.6 | #154 (Claude Opus 4 (20250514)) |
| 3 | Gemini 2.5 Pro (Preview 06-05) | 25.3 | #106 |
| 4 | Claude Sonnet 4 (20250514) | 25.1 | #213 |
| 5 | Claude 3.7 Sonnet (20250219) (Thinking) | 25.1 | #201 (Claude 3.7 Sonnet (20250219)) |
| 6 | Claude Sonnet 4 (20250514) (Thinking) | 24.3 | #213 (Claude Sonnet 4 (20250514)) |
| 7 | Claude 3.5 Sonnet (20241022) | 23 | #291 |
| 8 | Claude 3.7 Sonnet (20250219) | 22.5 | #201 |
| 9 | Claude 3.5 Sonnet (20240620) | 22 | #299 |
| 10 | GPT-4.1 | 21.1 | #248 |
| 11 | GPT-4.1 Mini | 20.8 | #355 |
| 12 | Gemini 2.5 Pro (Preview 05-06) | 20.8 | #147 |
| 13 | Doubao-1.5-Thinking-Pro | 20.1 | #134 |
| 14 | GPT-4o | 17.2 | #343 |
| 15 | DeepSeek R1 0528 | 17.1 | #222 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=web-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.