F5 Labs AI Security Leaderboard - CASI: leaderboard
F5 Labs' CASI (Comprehensive AI Security Index, from F5 AI Red-Team, formerly CalypsoAI) scores how well a model resists prompt-injection and jailbreak attacks drawn from a monthly refreshed attack library, weighting attacks by the sophistication needed to succeed. Scale 0-100, higher is more secure. F5 publishes only its ten highest-scoring models each month, so this board is a monthly top-10 snapshot.
Metric: CASI (Comprehensive AI Security Index), 0-100, higher is more secure (October 2026 top-10 snapshot). Source: www.f5.com. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 5 | 97.6 | #65 |
| 2 | GPT-5.4 | 95.76 | #83 |
| 3 | Claude Opus 5 | 94.75 | #16 |
| 4 | GPT-5.2 | 94.33 | #115 |
| 5 | GPT-5.4 Nano | 93.95 | #359 |
| 6 | GPT-6 Luna | 93.52 | #61 |
| 7 | Grok 4.7 (High) | 93.46 | #26 (Grok 4.7) |
| 8 | Nemotron 3 Ultra | 91.13 | #245 |
Interactive version: theaggregate.ai/benchmark?slug=f5-labs-ai-security-leaderboard-casi · How It Works · Data refreshed daily, snapshot 2026-11-05.