code benchmark
SecCodeBench
SecCodeBench evaluates LLM coding agents on secure code generation and vulnerability detection, testing the ability to produce code that is both functional and free from security vulnerabilities.
Updated Aug 7, 2026
Published models1
Registry coverage1
MetricScore
EvidenceB
SecCodeBench leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | AC | 68.3% | 100.0% | 1 | C |
SecCodeBench highlights
The top published results on this benchmark's own scale.
What is SecCodeBench?
Definition and scoring fields from the benchmark registry.
SecCodeBench evaluates LLM coding agents on secure code generation and vulnerability detection, testing the ability to produce code that is both functional and free from security vulnerabilities.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- SecCodeBench
- Modality
- text
- Primary category
- code
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- seccodebench|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about SecCodeBench.
Which model scores highest on SecCodeBench?
Qwen3.5-397B-A17B is currently ranked first with 68.3%.
What does SecCodeBench measure?
SecCodeBench evaluates LLM coding agents on secure code generation and vulnerability detection, testing the ability to produce code that is both functional and free from security vulnerabilities.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.