safety benchmark
Cybersecurity CTFs
Cybersecurity Capture the Flag (CTF) benchmark for evaluating LLMs in offensive security challenges. Contains diverse cybersecurity tasks including cryptography, web exploitation, binary analysis, and forensics to assess AI capabilities in cybersecurity problem-solving.
Updated Aug 11, 2026
Models3
Model coverage3
MetricScore
EvidenceB
Cybersecurity CTFs Ranking
Higher score ranks better on this benchmark.
| 01 | OP | 77.6% | 100.0% | 3 | C | |
| 02 | AN | 46.9% | 50.0% | 3 | C | |
| 03 | OP | 28.7% | 0.0% | 3 | C |
Cybersecurity CTFs Score Distribution
A closer view of the leading scores on this benchmark.
Cybersecurity CTFs Highlights
The leading models and scores on this benchmark.
What is Cybersecurity CTFs?
What Cybersecurity CTFs measures and how its scores work.
Cybersecurity Capture the Flag (CTF) benchmark for evaluating LLMs in offensive security challenges. Contains diverse cybersecurity tasks including cryptography, web exploitation, binary analysis, and forensics to assess AI capabilities in cybersecurity problem-solving.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- Cybersecurity CTFs
- Modality
- text
- Primary category
- safety
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- cybersecurity-ctfs|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about Cybersecurity CTFs.
Which model scores highest on Cybersecurity CTFs?
GPT-5.3 Codex is currently ranked first with 77.6%.
What does Cybersecurity CTFs measure?
Cybersecurity Capture the Flag (CTF) benchmark for evaluating LLMs in offensive security challenges. Contains diverse cybersecurity tasks including cryptography, web exploitation, binary analysis, and forensics to assess AI capabilities in cybersecurity problem-solving.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
3 model results are currently shown.
Does this benchmark affect the overall score?
Yes. This benchmark can contribute to the current LLMBoard capability score.