ExploitBench v8-bench (AutoNudge): leaderboard
Metric: Capability coverage (%), all-runs view with reported seed/turn budgets; AutoNudge. Source: exploitbench.ai. Saturation forecast: Around January 2027. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Mythos Preview | 78 |
| 2 | GPT-5.5 | 47 |
| 3 | Claude Opus 4.7 | 28 |
| 4 | Claude Sonnet 4.6 | 26 |
| 5 | Kimi K2.6 | 18 |
| 6 | GLM-5.1 | 18 |
| 7 | Gemini 3.1 Pro (Preview) | 16 |
| 8 | Claude Haiku 4.5 | 14 |
| 9 | MiniMax-M2.7 | 13 |
Interactive version: theaggregate.ai/benchmark?slug=exploitbench-v8-bench-autonudge · How It Works · Data refreshed daily, snapshot 2026-10-09.