LLM Stats (ExploitBench): leaderboard
ExploitBench as aggregated by LLM Stats — offensive-security agent tasks built from patched browser-engine vulnerabilities, graded on how far a model gets toward a working exploit.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Around January 2027. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra | 100 | #6 |
| 2 | Claude Fable 5 | 78 | #16 |
| 3 | GPT-5.6 Sol | 73.5 | #19 |
| 4 | GLM-5.3 | 54.4 | #39 |
| 5 | GPT-5.6 Terra | 52.9 | #41 |
| 6 | MiMo-V2.6-Pro | 47.9 | #42 |
| 7 | GPT-5.6 Luna | 33.2 | #68 |
| 8 | MiMo-V2.6-Flash | 25.3 | #70 |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-exploitbench · How It Works · Data refreshed daily, snapshot 2026-10-20.