Benchmark
Sayf SECURE MAET accuracy
Higher scores are better.
Benchmark sourceScores from Sayf cybersecurity evaluations data docs.
- Models measured
- 6
- Scores recorded
- 6
- Sources
- 1
- IQ weight
- 0.95
Score distribution
- 0.906
- Mean
- 0.923
- Median
- 0.028
- Std dev
- 0.873
- 10th percentile
- 0.941
- 90th percentile
- 0.867
- Min
This benchmark has a weight of 0.95 for chat models, measured over 6 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6secure_maet· Sayf cybersecurity evaluations data docs | 0.941 |
| 2 | GPT 5.4secure_maet· Sayf cybersecurity evaluations data docs | 0.931 |
| 3 | Gemma 4 31B ITsecure_maet· Sayf cybersecurity evaluations data docs | 0.923 |
| 4 | Qwen3.6 35B A3Bsecure_maet· Sayf cybersecurity evaluations data docs | 0.902 |
| 5 | GPT OSS 20bsecure_maet· Sayf cybersecurity evaluations data docs | 0.873 |
| 6 | Llama 3.3 70B Instructsecure_maet· Sayf cybersecurity evaluations data docs | 0.867 |
Data credits: Sayf-eval. Copyright (c) 2026 Qatar Computing Research Institute (QCRI). MIT License. (licence)