Which AI model scores highest on CWE-bench v1?

Gemini 4 Carbon by Google holds the best CWE-bench v1 result among tracked models, at 71% (released Oct 28 2026). Higher scores are better on this benchmark.

What are the top 1 models on CWE-bench v1?

1. Gemini 4 Carbon (Google) — 71%.

How many models have a published CWE-bench v1 score?

1 tracked model has a published CWE-bench v1 score. Scores are collected from published lab reports and benchmark sources, with recorded sources credited alongside the results. We do not re-run the evaluations.

Security patching

CWE-bench v1

Collinear AI’s programmatically graded security-patch pass@1 rate; separate from exploit-generation benchmarks

Rankings

Higher is better

Gemini 4 Carbon

71%

CWE-bench v1 — frequently asked questions

: Collinear AI’s programmatically graded security-patch pass@1 rate; separate from exploit-generation benchmarks
: Gemini 4 Carbon by Google holds the best CWE-bench v1 result among tracked models, at 71% (released Oct 28 2026). Higher scores are better on this benchmark.
: 1. Gemini 4 Carbon (Google) — 71%.
: 1 tracked model has a published CWE-bench v1 score. Scores are collected from published lab reports and benchmark sources, with recorded sources credited alongside the results. We do not re-run the evaluations.