Google Unveils Affordable Alternative to Massive AI Security Models with Mythos Competitor

Google Introduces Gemini 3.6 Flash and Innovative Security Model

Google is set to launch Gemini 3.6 Flash, along with a cutting-edge security model designed to swiftly detect and address security vulnerabilities. This new development aims to enhance cybersecurity measures and provide a more efficient and cost-effective solution for organizations.

Enhanced Security with Gemini 3.5 Flash Cyber

Google’s cybersecurity model, built on the foundation of Gemini 3.5 Flash, will initially be accessible to governments and trusted partners through CodeMender, Google’s security-focused coding agent. This model, known as 3.5 Flash Cyber, offers rapid scanning capabilities at a low cost, enabling AI agents to identify vulnerabilities efficiently.

Advancements in Gemini 3.6 Flash and Flash-Lite

Google’s upgraded Gemini 3.6 Flash introduces improvements in coding and multimodal performance, while the new 3.5 Flash-Lite model is highlighted as the most cost-effective option within the 3.5 series.

Performance Comparison and Unique Discoveries

According to Google, 3.5 Flash Cyber has demonstrated competitive performance when compared to larger AI models on the CyberGym AI cybersecurity benchmark. The model identified 55 unique confirmed issues in the V8 JavaScript Engine, surpassing the findings of other models. Notably, 3.5 Flash Cyber uncovered 10 issues that were previously undetected by other models, showcasing its efficacy in identifying new vulnerabilities.

Update, July 21st: Added information about Google’s other Gemini models.