CVE-Bench: leaderboard
Cybersecurity benchmark for autonomous web vulnerability exploitation across 40 critical CVEs in zero-day and one-day settings.
Metric: pass@1 (%). Source: github.com. Source: deploymentsafety.openai.com. Saturation forecast: Estimated already saturated. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 | 93.1 | #31 |
| 2 | GPT-5.5 Pro | 93.1 | #17 |
| 3 | GPT-5.3 Codex | 90.2 | #69 |
| 4 | GPT-5.4 (Thinking) | 86.27 | #79 (GPT-5.4) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=cve-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.