ExploitGym (version unspecified; 2h timeout cutoff): leaderboard
Metric: Successful intended exploits (#); version unspecified; 2h timeout cutoff. Source: www.cybergym.io. Saturation forecast: Around 2028. 3 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Mythos 5 | 181 |
| 2 | Claude Opus 5 | 171 |
| 3 | Claude Opus 4.8 | 80 |
Interactive version: theaggregate.ai/benchmark?slug=exploitgym-version-unspecified-2h-timeout-cutoff · How It Works · Data refreshed daily, snapshot 2026-10-09.