Published 11 hours ago • loading... • Updated 2 hours ago
Google Announces Gemini 3.6 Flash and Cybersecurity AI, Teases 3.5 Pro and Gemini 4
Google says the new models cut token use, lower costs and add a cybersecurity option limited to governments and trusted partners.
On Tuesday, July 21, 2026, Google released three new Gemini models: 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber, designed to provide "efficiency, low latency, and reliability" for developers running AI agents at scale.
These releases prioritize cost-efficiency as companies face rising token budgets. Google CEO Sundar Pichai stated earlier this year that using a mix of Flash models "could save firms over $1 billion a year."
The "workhorse" Gemini 3.6 Flash model improves coding performance while using about 17% fewer output tokens than its predecessor. Gemini 3.5 Flash-Lite offers high-throughput processing at $0.3 per million input tokens.
For cybersecurity, 3.5 Flash Cyber detects and patches software vulnerabilities within Google's CodeMender agent. The model remains in a limited-access pilot exclusively for governments and trusted partners due to dual-use safety risks.
Google confirmed Gemini 3.5 Pro remains in testing with partners, available "as soon as it's ready." The company has begun its "most ambitious pre-training run yet" for Gemini 4, signaling continued development ahead.