dfs-large1 Matches Frontier Models in Vulnerability Discovery Using Open Weights

jfiance · x · 2026-07-31

DepthFirst announced dfs-large1, their newest cybersecurity model achieving best-in-class performance on vulnerability detection tasks. Besides frontier AI labs, only a handful of companies have built specialized models reaching state-of-the-art in their domain.

Built on the open-source GLM-5.2 model, dfs-large1 was post-trained with reinforcement learning inside DepthFirst's security infrastructure. It achieves top-tier performance on their benchmark for long-horizon vulnerability discovery across complex repositories, proving that American companies can build best-in-class domain models using open weights.

Related event: DepthFirst Launches GLM-5.2-based Cybersecurity Model dfs-large1(3 posts)→

Original post →

More from Models

Models channel →