Google Releases New Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber Models

Google is making three new Gemini AI models available today to make up for the delayed release of its Gemini 3.5 Pro model, which was announced at its I/O conference in May. The headliner is the Gemini 3.6 Flash model, but there’s also the fastest and more cost-effective 3.5 Flash-Lite and the cybersecurity-focused 3.5 Flash Cyber.

According to Google, Gemini 3.6 Flash is more efficient and cheaper to use than 3.5 Flash for agentic tasks. The new model also comes with additional safeguards in the domains of Chemical, Biological, Radiological, and Nuclear (CBRN) and cyber offense misuses.

“3.6 Flash not only delivers a step up in coding and knowledge work, but it does this while meaningfully improving token efficiency,” the company said. “It also takes fewer reasoning steps and tool calls to accomplish multi-step workflows.”

Google’s new Gemini 3.5 Flash-Lite model is optimized for agentic workflows, and it’s now the fastest model in the 3.5 series of Gemini models. It’s also a cost-efficient model with “a strong price-to-performance ratio for developers and customers running high throughput production traffic,” the company explained.

Lastly, Gemini 3.5 Flash Cyber is Google’s response to the latest flagship cybersecurity models from rivals like OpenAI and Anthropic. Google said that it can find and fix cybersecurity vulnerabilities at “a lower price per token than larger models.”

At the moment, access to Gemini 3.5 Flash Cyber is restricted to governments and trusted partners via a limited-access pilot program. Gemini 3.6 Flash and 3.5 Flash-Lite, however, are rolling out to the Gemini app, Google Search,  Google AI Studio, Android Studio, and the Gemini Enterprise app.