Google's Gemini 3.6 Flash AI model series is here
What's the story
Google DeepMind has launched three new models in its Gemini series: the 3.6 Flash, 3.5 Flash-Lite, and the specialized 3.5 Flash Cyber. The company describes the Gemini 3.6 Flash as its "workhorse model," promising improved performance in coding, knowledge work, and multimodal tasks while cutting token usage by up to 17%. This makes it more cost-effective than its predecessor, the Gemini 3.5 Flash.
Model details
'Flash Cyber' model to be available for governments
The Gemini 3.5 Flash-Lite is the most affordable model in this series, while the Gemini 3.5 Flash Cyber is a specialized version designed to detect and patch cybersecurity vulnerabilities at a reasonable price point.
However, this model will only be available to governments and trusted partners through a limited access pilot program, as per Google's announcement.
Race
Google's race against OpenAI, Anthropic
The launch of these new models is part of Google's strategy to provide efficiency, latency, and reliability to customers building AI agents at scale.
However, the update doesn't include the much-anticipated upgrade to Google's flagship model, Gemini Pro.
This model was last updated in February and its absence from this release highlights the intense competition in the AI space, with rival labs like OpenAI and Anthropic releasing new models at a fast pace.
Pro delay
Gemini 3.5 Pro expected to launch soon
The launch of the highly-anticipated Gemini 3.5 Pro has been delayed due to coding performance issues, with no new date announced yet.
Despite this setback, Google DeepMind product lead Logan Kilpatrick said that the company is currently testing Gemini 3.5 Pro with partners and hopes to "land soon."
He also revealed that they have begun their most ambitious pre-training run yet for Gemini 4.