PawBench - QwenPaw — leaderboard
Metric: Overall Score (%). Source: agentscope-ai.github.io. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.6 Max Preview | 78.35 |
| 2 | Claude Opus 4.6 | 78.27 |
| 3 | Qwen 3.7 Max | 77.65 |
| 4 | DeepSeek V4 Pro | 75.64 |
| 5 | Qwen 3.6 Plus | 75.01 |
| 6 | Qwen 3.6 27B | 72.66 |
| 7 | GLM-5.1 | 71.07 |
| 8 | Qwen 3.6 35B A3B | 68.28 |
| 9 | Kimi K2.6 | 66.59 |
Interactive version: aibenchmarks.dev/benchmark?slug=pawbench-qwenpaw · How the rankings work · Data refreshed daily, snapshot 2026-07-19.