Introducing Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

Gemini 3.6 Flash is priced at $1.50/$7.50 per million tokens and is said to use 17% fewer output tokens than 3.5 Flash. Flash-Lite is priced at $0.30/$2.50 and is reported at 350 output tokens per second, with gains on Terminal-Bench, GDPval and SWE-Bench Pro.

3.5 Flash Cyber is a fine-tune for finding and fixing vulnerabilities inside CodeMender and will be limited to governments and trusted partners. Gemini 3.5 Pro is still in partner testing; Gemini 4 pre-training has started.

This is Google DeepMind’s own model announcement, so prices, benchmark deltas and safety claims are vendor-reported. The more important move is product segmentation: efficiency, throughput and restricted cyber capability are now separate SKUs.

Signal Desk's read: token efficiency is the real 3.6 Flash story. A cheaper, less verbose workhorse can change agent unit economics more than a modest accuracy bump.

The cyber model’s limited access is the other signal. Google is treating offensive-adjacent capability as a distribution problem, not only a training problem.

3.6 Flash and Flash-Lite are available in the Gemini API, AI Studio, Gemini Enterprise and the Gemini app; Flash-Lite is also rolling into Search. Computer use is a built-in client-side tool.