Atlas

Benchmarks

← All benchmarks

CWE-bench v1: Software Supply Chain Failures

Code · 2026-09-25

Programmatic pass@1 on the OWASP 2025 A03 subset of CWE-bench v1; exact category denominator is not published.

Top models (higher is better)

ModelScore
Gemini 4 Argon75.0
Claude Fable 5.165.0
Claude Opus 5.565.0
Grok 4.763.0
DeepSeek V4.1 Flash60.0
Muse Spark 1.360.0
GPT-6 Astra58.0
GPT-6 Sol58.0
Grok 4.658.0
Hy4 Preview50.0
Inkling43.0
Loading Atlas data…