Overview
- Google and DeepMind announced three new models—Gemini 3.6 Flash, Gemini 3.5 Flash‑Lite, and Gemini 3.5 Flash Cyber—positioning them for higher capability, high‑throughput agent work, and security tasks respectively.
- Gemini 3.6 Flash and 3.5 Flash‑Lite are available to users and developers in Google’s tools and the Gemini app while Gemini 3.5 Flash Cyber will be access‑restricted to governments and trusted partners.
- Google claims the Flash series reduces token use by about 17% versus the prior generation and offers lower per‑token pricing to cut the operating cost of production agents and bulk workloads.
- Independent early tests published after the launch report notable regressions in Gemini 3.6 Flash, including broken frontend UI code generation, poor spatial reasoning, and apparent knowledge cutoff issues that undercut some developer use cases.
- Google says Gemini 3.5 Pro remains delayed and it has begun large‑scale pretraining for Gemini 4; the Cyber model reportedly found 55 issues in the V8 JavaScript engine, including 10 not previously detected, which shaped the decision to restrict access.