QwenWebBench
Can the model complete substantial programming work under this benchmark's agent setup?
Top rankings
One row per model, using its best reported score across effort settings.
| Rank | Model | Best score | Best reported setting |
|---|---|---|---|
| 1 | Claude Opus 4.5Anthropic | 1536 | Reported configurationQwen report1 value · 1 reportApr 21, 2026 · source |
| 2 | Qwen3.6-27BQwen | 1487 | Reported configurationQwen report1 value · 1 reportApr 21, 2026 · source |
| 3 | Qwen3.6-35B-A3BQwen | 1397 | Reported configurationQwen report2 values · 2 reportsApr 21, 2026 · source |
| 4 | Gemma4-31BGoogle | 1197 | Reported configurationQwen report2 values · 2 reportsApr 21, 2026 · source |
| 5 | Qwen3.5-397B-A17BQwen | 1186 | Reported configurationQwen report1 value · 1 reportApr 21, 2026 · source |
| 6 | Gemma4-26B-A4BGoogle | 1178 | Reported configurationQwen report1 value · 1 reportApr 15, 2026 · source |
| 7 | Qwen3.5-27BQwen | 1068 | Reported configurationQwen report2 values · 2 reportsApr 21, 2026 · source |
| 8 | Qwen3.5-35B-A3BQwen | 978 | Reported configurationQwen report1 value · 1 reportApr 15, 2026 · source |
Effort curve
Every sourced cost-linked effort value for this exact version. Lines connect complete sweeps only.
Definition and comparison boundaryinternal methodology
Can the model complete substantial programming work under this benchmark's agent setup?
Software implementation, debugging, or repository work under the published evaluation protocol. Higher is better. This benchmark reports points rather than percent correct.
This is a publisher-defined internal evaluation. The task set or grading details are not fully public, so treat it as directional evidence.
Internal bilingual front-end code-generation benchmark with seven categories, auto-render plus multimodal judge, and BT/Elo rating.
Method / source