The useful question is what changes for users, developers or buyers, and whether the announcement stays industry context or becomes something people can actually use.

Google announced three new models in the Gemini series: 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber, designed to enhance efficiency, latency, and reliability for AI agents. The updates respond to developer feedback, focusing on token efficiency and reduced costs. 3.6 Flash, priced at $1.50 per 1M input tokens and $7.50 per 1M output tokens, consumes 17% fewer output tokens than its predecessor while improving performance in coding and knowledge tasks.

3.5 Flash-Lite targets low-latency and high-throughput workflows, priced at $0.30 per 1M input tokens and $2.50 per 1M output tokens. It achieves 350 output tokens per second and outperforms earlier versions in coding, agentic tasks, and long-context evaluations. The model supports configurable thinking levels and includes built-in computer use tools for agentic systems.

3.5 Flash Cyber incorporates enhanced safety safeguards in domains such as Chemical, Biological, Radiological, and Nuclear (CBRN) and cyber offense misuses, reducing jailbreak susceptibility while minimizing refusals for beneficial uses. The model is positioned for secure deployment in high-risk scenarios.

The new models are available alongside ongoing development of Gemini 3.5 Pro and preparations for the next-generation Gemini 4. Early customer feedback highlights the improved cost-to-performance ratio and scalability of the Flash series for production AI agents.