Google DeepMind released Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber on Tuesday. The flagship 3.6 Flash variant claims improved coding and knowledge work capabilities while reducing token usage by up to 17% compared to the previous 3.5 Flash version. The 3.5 Flash-Lite model targets cost efficiency, whereas the new 3.5 Flash Cyber model is fine-tuned for identifying and fixing security vulnerabilities. This cybersecurity tool will be available exclusively to governments and trusted partners through a limited pilot program. The company states the focus is on delivering efficiency, latency, and reliability for customers building AI agents at scale.
The launch is notable for what it omitted alongside the new releases. There is no update to Gemini Pro, the model intended for complex reasoning, which last saw an update in February. While rivals like OpenAI and Anthropic have recently launched multiple frontier models, Google has not followed suit with a Pro update. Internal delays reportedly prevented the release of the 3.5 Pro despite earlier promises made in May. Product lead Logan Kilpatrick confirmed the 3.5 Pro is currently being tested with partners and is expected to launch soon. The team has also begun its most ambitious pre-training run yet for Gemini 4.
- Gemini 3.6 Flash reduces token costs by up to 17%
- Gemini 3.5 Flash Cyber is restricted to government and trusted partners
- Gemini Pro 3.5 launch faces internal delays pending performance goals
