Enkrypt AI - Insecure Code Risk: leaderboard
Enkrypt AI’s LLM safety leaderboard: the share of generated code carrying an exploitable weakness. 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. 300 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.2 | 0 | #110 |
| 2 | Claude Fable 5 | 0 | #16 |
| 3 | Claude Opus 5 | 0.44 | #12 |
| 4 | Grok 4.5 | 0.44 | #48 |
| 5 | Claude Opus 5.5 | 0.44 | #3 |
| 6 | Grok 4.6 | 0.44 | #26 |
| 7 | MiniMax-M2.7 | 0.89 | #258 |
| 8 | GPT-OSS-120B | 1.33 | #341 |
| 9 | GPT-5 Nano | 1.33 | #425 |
| 10 | Mercury 2 | 1.33 | #433 |
| 11 | GPT-5 | 2.22 | #98 |
| 12 | Inkling | 2.22 | #126 |
| 13 | Grok 4.7 | 2.67 | #25 |
| 14 | GPT-5 Mini | 3.11 | #186 |
| 15 | QwQ 32B-Preview | 3.11 | #651 |
No result here: #6 GPT-6 Astra, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=enkrypt-ai-insecure-code-risk · How It Works · Data refreshed daily, snapshot 2026-10-20.