Google launches Gemini 3.8 Flash Cyber, a frontier cybersecurity model that finds and patches vulnerabilities autonomously
OfficialLoganK · x · 2026-09-10
Google DeepMind released Gemini 3.8 Flash Cyber, its most capable cybersecurity model built for defenders. It hits frontier-level performance on the CyberGym vulnerability-discovery benchmark (Pass@1, surpassing 3.5 Flash Cyber), can autonomously navigate codebases across 20 programming languages to uncover critical hidden vulnerabilities, and auto-generates validated patch code. It's already used extensively inside Google by Chrome, Wiz, and Cloud, with external expansion coming. A new Fairwind Program gives trusted partners early access to defend against AI-driven cyber threats.
Related event: Google Unveils Gemini 3.8 Flash Cyber Security Model(2 posts)→
More from Models
- Nanbeige4.3-33B benchmarked: 59 tok/s single stream, ~1,060 tok/s across 64 concurrent runs — MaziyarPanahi · 2026-10-26
- Only 2 of 25 full ChatGPT rewrites flagged 100% AI by detectors — paulnovosad · 2026-10-26
- Qwen 4 Reportedly Spotted in the Arena, Hints at Alibaba's Next-Gen Model Testing — AIFlow_ML · 2026-10-26
- Nathan Lambert's first open models lecture: distillation, risks and the ecosystem explained — natolambert · 2026-10-26
- Nanbeige 4.3-33B flops real-world test: zero compilable output vs Qwen — julianharris · 2026-10-26
- Claude's safety classifier blocks SS7 defense research, pushing a security dev to an abliterated GLM — l4rz · 2026-10-26