Agent Security League - Security Correctness — leaderboard
Metric: Security Correctness (%). Source: www.endorlabs.com. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5 | 29 |
| 2 | GPT-5.5 | 24 |
| 3 | GPT-5.6 Sol | 23.5 |
| 4 | GPT-5.4 | 21.8 |
| 5 | Claude Opus 4.8 | 20.7 |
| 6 | Claude Sonnet 5 | 19.6 |
| 7 | Claude Opus 4.7 | 18.4 |
| 8 | Gemini 3.5 Flash | 17.9 |
| 9 | Gemini 3.1 Pro (Preview) | 15.1 |
| 10 | Claude Sonnet 4.6 | 12.3 |
| 11 | GLM-5.2 | 12 |
| 12 | Claude Opus 4.6 | 11.2 |
| 13 | Gemini 3 Pro | 9.5 |
| 14 | Claude Opus 4.5 | 8.4 |
| 15 | Claude Sonnet 4 | 7.8 |
Interactive version: theaggregate.ai/benchmark?slug=agent-security-league-security-correctness · How the rankings work · Data refreshed daily, snapshot 2026-07-22.