Artificial Analysis Cyber Index

Artificial Analysis' composite index of cyber-defense capability: finding vulnerabilities in real codebases, reproducing and validating them, and patching them without breaking existing functionality. It averages CWE-Bench-AA, DeepsecBench-AA and CyberGym-E2E-AA.

ModelsModels with a result.19
Top resultBest result on this benchmark.64.1%Kimi K3.1 Pro
Top-3 spreadPoints from first to third.8.0 pts
YearYear of release.2026

Result and price

17 of 19 results

Results

19 results
#ModelResultThe score from the source.
01Kimi K3.1 ProAbove fitFar above this model's index.64.1%
02Grok 4.756.4%
04GPT-6 Luna52.7%
05Mistral Large 4Above fitFar above this model's index.50.9%
08Kimi K341.3%
10GPT-6 Sol36.8%
11GLM-5.336.4%
17MiniMax M322.4%