Enkrypt AI - Jailbreak Risk: leaderboard
Enkrypt AI’s LLM safety leaderboard: how often a jailbreak attack gets the model past its own policy. It is reported as a risk, so lower is better here and the leader is the least risky model rather than the most capable one.
Metric: Risk Score. Source: leaderboard.enkryptai.com. Saturation forecast: Estimated already saturated. 298 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5 | 0 | #16 |
| 2 | Claude 3.5 Sonnet (20241022) | 0.2 | #291 |
| 3 | Claude 3.7 Sonnet (20250219) | 0.72 | #201 |
| 4 | Inkling | 0.72 | #126 |
| 5 | O1 Preview | 0.86 | #193 |
| 6 | O1 | 0.92 | #183 |
| 7 | Claude 3.5 Haiku (20241022) | 1.12 | #588 |
| 8 | Muse Spark 1.1 | 1.44 | #52 |
| 9 | Claude 3 Opus (20240229) | 1.51 | #412 |
| 10 | Claude Opus 4.7 | 1.73 | #49 |
| 11 | GPT-5 Nano | 1.86 | #425 |
| 12 | Hy3 | 1.87 | #93 |
| 13 | Claude 3 Sonnet (20240229) | 1.97 | #677 |
| 14 | Claude Opus 4 (20250514) | 1.97 | #154 |
| 15 | Claude Sonnet 5 | 2.01 | #57 |
No result here: #6 GPT-6 Astra, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=enkrypt-ai-jailbreak-risk · How It Works · Data refreshed daily, snapshot 2026-10-20.