Cantina releases apex-flash-1, an open-weights security model post-trained on real paid vulnerabilities
ctjlewis · x · 2026-10-02
Security firm Cantina, with Yeta Labs, released apex-flash-1, its first open-weights model for security research, post-trained on real vulnerabilities found via its paid bounty programs. An abliterated variant with fewer refusals is also available for authorized research workflows.
The team argues the dominant safety debate—controlling advanced cyber capabilities via proprietary APIs and usage restrictions—is flawed because cybersecurity is adversarial: attackers ignore acceptable-use policies and procurement rules, and capability will spread through open weights anyway. The linked post digs into governance questions around who should access such capabilities.
More from Models
- DeepMind Senior Staff Engineer Calls Bloomberg's Gemini 4 Employee Gripe Report "BS" — MrZiruiWang · 2026-10-02
- Local benchmark: community 9B model shows 4x lower latency on RTX 5070 vs peers — Storge2 · 2026-10-02
- Fable 5.5 release 'impending, as soon as next week', claims Bindu Reddy — bindureddy · 2026-10-02
- Open weights will be good enough: why frontier AI smarts won't matter for daily life — Dan_Jeffries1 · 2026-10-02
- Open-weight models trail frontier by just 4 months — here's when to use them — TechPreacher · 2026-10-02
- MiniMax M3.1 Flash joins RSI Arena mid-run as compute competition heats up — my_cat_can_code · 2026-10-02