Z.ai GLM-5.3 Launch - ExploitGym (6h): leaderboard
ExploitGym real-world exploit-generation tasks completed under a six-hour throughput-normalized budget, as reported in Z.ai's GLM-5.3 launch table; Z.ai-run frontier comparison.
Metric: Successful Intended Exploits (#). Source: z.ai. Saturation forecast: Around December 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol | 293 | #19 |
| 2 | Claude Fable 5 | 247 | #16 |
| 3 | GLM-5.3 | 130 | #39 |
| 4 | Claude Opus 4.8 | 120 | #37 |
| 5 | Kimi K3 | 70 | #30 |
| 6 | GLM-5.2 | 39 | #73 |
| 7 | Qwen 3.8 Max | 26 | #35 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=z-ai-glm-5-3-launch-exploitgym-6h · How It Works · Data refreshed daily, snapshot 2026-10-20.