Google announced Gemini 3.5 Flash Cyber, an AI model focused on identifying and patching security vulnerabilities. According to a company blog post, Google positions the model as a “cost-efficient and highly capable alternative” to larger, more expensive AI systems, including Anthropic’s Mythos. The model is built on Gemini 3.5 Flash and will initially be offered to governments and trusted partners through CodeMender, Google’s security-focused coding agent.
Google says CodeMender can invoke Gemini 3.5 Flash Cyber repeatedly at high speed and low cost, enabling its AI agents to scan more code paths and surface additional vulnerabilities.
Why it matters
By offering a lower-cost option, Google targets the expense associated with running large AI security models, potentially making automated vulnerability detection more accessible to its early partners.
Who should care
Governments and trusted partners with early access, along with organizations evaluating AI-assisted approaches to code security, are the primary audiences for this launch.