Google launches Gemini 3.6 Flash in push for faster, cheaper AI models
By Suad Seferi ·
Google has expanded its Gemini lineup with new models designed to improve speed, efficiency and real-world AI performance. The update introduces Gemini 3.6 Flash, Gemini 3.5 Flash-Lite and Gemini 3.5 Flash Cyber, marking another step in the company’s move toward more capable agentic AI systems. Gemini 3.6 Flash is the main release in the update. Google says the model is built for everyday tasks, coding, multi-step workflows and agentic execution, while also improving token efficiency and lowering cost compared with the previous Flash generation. The company also says Gemini 3.6 Flash is optimized for tasks such as code generation, code refactoring, reasoning and computer use. Google’s developer documentation says the model supports text, image, video, audio and PDF inputs, along with features such as function calling, code execution, file search, structured outputs and search grounding. Alongside that release, Google introduced Gemini 3.5 Flash-Lite, a lighter model aimed at low-latency and high-throughput use cases. It is now available in the Gemini app and is also rolling out in Google Search, giving users a faster option for simpler tasks and agentic workflows. The third model, Gemini 3.5 Flash Cyber, is designed for cybersecurity use cases, especially finding and fixing vulnerabilities at scale. Google says access to that model will be limited to governments and trusted partners under a pilot program. The update shows how Google is shaping Gemini around practical use rather than raw model size alone. In a market where speed, cost and reliability matter more every day, the company is clearly trying to position Gemini as a serious platform for developers, enterprises and AI agents.