ZClawBench: leaderboard
ZClawBench evaluates Claw-style agent task execution quality, measuring a model's ability to autonomously complete complex multi-step coding tasks in real-world environments.
Metric: Score (%). Source: huggingface.co. Source: benchmarklist.com. Saturation forecast: Around January 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 | 65.4 | #63 |
| 2 | GLM-5-Turbo | 56.4 | #119 |
| 3 | Gemini 3.1 Pro (Preview) | 50.3 | #58 |
| 4 | GLM-5 | 48.2 | #141 |
| 5 | Kimi K2.5 | 40.2 | #144 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=zclawbench · How It Works · Data refreshed daily, snapshot 2026-10-20.