Google Launches Gemini 3.8 Flash AI Models With Cybersecurity Variant

Google has launched Gemini 3.8 Flash, the latest in its series of AI models, alongside a specialized cybersecurity variant called Gemini 3.8 Flash Cyber. The release, announced on September 2, 2026, marks the company's third Flash-tier model in six weeks, with both variants built on the same underlying architecture but tuned for different purposes. The general-purpose Gemini 3.8 Flash is designed for developers building agentic applications and workflows. It is also available to Google AI Pro and Ultra subscribers through the Gemini app. Google states the model "works harder" by taking additional reasoning steps on complex tasks, which can increase token usage but improves performance. On the DeepSWE v1.1 benchmark for autonomous engineering problem-solving, the model scored 73.7, outperforming most larger frontier models. It also showed gains over its predecessors in areas like financial analysis, legal workflows, and long video understanding. Pricing for Gemini 3.8 Flash is set at an introductory rate of $0.75 per million input tokens and $3.75 per million output tokens. These prices are expected to rise to $1.50 and $7.