← Back to Model Beat
Research·Oct 5·all news from October 5, 2026

Can an Open Model Do Security Research? Cantina’s apex-flash-1 Solves 40 of 60 Held-Out Bug Tasks

Cantina Security and Yeta Labs have released apex-flash-1, an open-weights model fine-tuned for vulnerability research using the GLM-5.3-Flash architecture. In testing, the model successfully identified 40 out of 60 held-out bug tasks. Because it is released under an MIT license, the model can be deployed locally on standard infrastructure, offering researchers an accessible tool for automated security analysis.

Covered by 1 source

Related stories

ResearchAI Whistleblowers, Google, OpenAI, Meta to Face New York City CouncilOct 4 · 12 sourcesResearchOpenAI's safety crisis keeps getting worse and the company keeps making it worseOct 8 · 10 sourcesResearchDisrupting AI-enabled “false front” operationsOct 8 · 2 sourcesResearchAI Solves a Major Unsolved Math Problem. Not Everyone Is HappyOct 4 · 4 sources