PawBench - Hermes — leaderboard
Metric: Overall Score (%). Source: agentscope-ai.github.io. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 | 78.44 |
| 2 | Qwen 3.7 Max | 72.29 |
| 3 | DeepSeek V4 Pro | 72.06 |
| 4 | Qwen 3.6 Plus | 70.41 |
| 5 | Qwen 3.6 27B | 68.2 |
| 6 | Qwen 3.6 Max Preview | 68.09 |
| 7 | Kimi K2.6 | 66.36 |
| 8 | GLM-5.1 | 63.19 |
| 9 | Qwen 3.6 35B A3B | 56.74 |
Interactive version: aibenchmarks.dev/benchmark?slug=pawbench-hermes · How the rankings work · Data refreshed daily, snapshot 2026-07-19.