CyberMetric: leaderboard
Multiple-choice cybersecurity knowledge Q&A across 500 questions covering threats, vulnerabilities, defenses, and best practices. Tests breadth of security knowledge.
Metric: 500-question accuracy (%). Source: arxiv.org. Saturation forecast: Estimated already saturated. 25 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o | 93.4 | #343 |
| 2 | GPT-4 Turbo | 93.3 | #404 |
| 3 | Mixtral 8x7B Instruct | 91.8 | #1170 |
| 4 | GPT-3.5 Turbo | 87.3 | #865 |
| 5 | Gemini 1.0 Pro | 85.05 | #933 |
| 6 | Mistral 7B Instruct (v0.2) | 78.4 | #1333 |
| 7 | zephyr-7B-beta | 76.4 | #1330 |
| 8 | Llama 3 8B Instruct | 76.2 | #1129 |
| 9 | Gemma 1.1 7B (IT) | 75.4 | #1368 |
| 10 | Llama 2 70B | 73.4 | #1201 |
| 11 | Qwen1.5-MoE-A2.7B | 64.6 | #1373 |
| 12 | Qwen 1.5 7B | 60.6 | #1327 |
| 13 | Qwen-7B | 58 | #1402 |
| 14 | Llama3-ChatQA-1.5-8B | 52.8 | #1364 |
| 15 | Phi-2 | 48 | #1511 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=cybermetric · How It Works · Data refreshed daily, snapshot 2026-10-20.