The idea of continuous innovation, where AI companies build on the products they are pushing out, updating alongside the pace of innovation, is a major one in the AI Era. One of the clearest examples of this is the evolution of AI models.
Now, Google has unveiled three new Gemini models: Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber, with a focus on greater efficiency, lower costs, and scaling AI agents. Here’s what you need to know.
Gemini 3.6 Flash
Gemini 3.6 Flash will replace Gemini 3.5 Flash as Google’s primary AI workhorse model for everyday tasks. The new model has better capacity for coding, reasoning, and advanced multimodal capabilities. In terms of efficiency and cost-effectiveness, it uses around 17% fewer output tokens.
Gemini 3.6 Flash supports text, images, video, audio, and PDFs with a context window of over one million tokens and has been built to enable AI agents to perform multiple tasks efficiently and with better accuracy.
Community Summit North America is the largest independent innovation, education, and training event for Microsoft business applications delivered by Expert Users, Microsoft Leaders, MVPs, and Partners. Register now to attend Community Summit in Nashville, TN from October 11-15.
Gemini 3.5 Flash-Lite
Gemini 3.5 Flash-Lite is now Google’s most cost-effective Gemini model. It’s designed for high-volume automation and “built to scale agentic workflows.” Google says the new model benefits from incredibly low latency, is the fastest model in the 3.5 series, and is an ideal fit for chatbots, customer service, and repetitive AI tasks.
Gemini 3.5 Flash Cyber
Perhaps the most innovative model in the new series is Gemini 3.5 Flash Cyber. It is Google’s first purpose-built, cybersecurity-focused Gemini model, specifically fine-tuned to identify and fix software vulnerabilities.
During this initial release wave, the model will only be available to governments and trusted partners through a limited-access program. However, we will be keeping a close eye on its performance, and you can expect to hear more about how the model behaves in the field.
We can’t discuss this announcement without mentioning the elephant in the room: the absence of Gemini 3.5 Pro, which was highly anticipated. However, even though Google hasn’t yet released its follow-up flagship model, there is still a lot for customers to be excited about with these three new models.
These models will help businesses deploying AI at scale to lower their operating costs. Customers can expect faster response times for AI agents, and from a developer’s perspective, improved coding performance.
Final Thought
Google’s latest Gemini releases, which notably did not include Gemini 3.5 Pro, really show that the AI race today is not solely about building the largest or most powerful models. Instead, it’s focused on delivering faster, more efficient, and specialized AI — case in point: the first Gemini model designed specifically for cybersecurity — that customers can deploy at scale.
Gemini 3.6 Flash, Flash-Lite, and Flash Cyber show that Google is making incremental advancements in power and performance. The company is heavily investing in these small, practical improvements that are crucial for enterprise customers.
Community Summit North America is the largest independent innovation, education, and training event for Microsoft business applications delivered by Expert Users, Microsoft Leaders, MVPs, and Partners. Register now to attend Community Summit in Nashville, TN from October 11-15.
