Google is rolling out a new cybersecurity-focused AI model called Gemini 3.5 Flash Cyber, positioning it as a lower-cost option for finding and fixing software vulnerabilities. The launch points to growing demand for AI security tools that can handle practical coding and defense tasks without the expense typically associated with larger models.
According to the announcement, Gemini 3.5 Flash Cyber is designed to identify weaknesses in code and help patch them. That puts the model in direct competition with bigger AI security offerings, including systems described as large security models such as Mythos, while emphasizing efficiency and affordability.
Google is also integrating Gemini 3.5 Flash into CodeMender, where it can be used to detect and remediate vulnerabilities as part of development workflows. This kind of integration suggests the company is aiming at teams that want AI assistance embedded into existing coding and security processes rather than used as a standalone tool.
The release highlights a broader shift in the AI market: companies are increasingly looking for specialized models that deliver focused performance at a more manageable price. With Gemini 3.5 Flash Cyber, Google appears to be targeting organizations that need faster, cheaper security support for code review and vulnerability patching.
