Google Releases Gemini 3.6 Flash and Specialized Security Models

Google continues to iterate on its Gemini series, moving from the 3.5 series to the 3.6 architecture while simultaneously developing specialized variants for security and efficiency. The company has also initiated the pre-training phase for its next-generation model, Gemini 4.

1. Gemini 3.6 Flash improves efficiency by consuming 17% fewer output tokens than the previous 3.5 Flash model. This reduction aims to lower costs for users, with pricing set at $1.50 per million input tokens and $7.50 per million output tokens.

2. The 3.5 Flash-Lite model provides high-speed performance, operating at 350 output tokens per second. It is priced at $0.30 per million input tokens and $2.50 per million output tokens, offering a lower-cost alternative for high-volume tasks.

3. Gemini 3.6 Flash includes enhanced safety safeguards. These protections target potential misuse related to chemical, biological, radiological, and nuclear (CBRN) threats and cyber offense activities.

4. Gemini 3.5 Flash Cyber demonstrates high efficacy in software security. In tests on the V8 JavaScript Engine, the model identified 55 unique vulnerabilities, outperforming the 47 identified by Gemini 3.5 Flash and the 36 found by Claude Opus 4.6.

1. The use of AI for vulnerability scanning remains a point of contention in the industry. Google DeepMind noted that some competitor models released after Opus 4.6 refuse to perform vulnerability scanning tasks entirely due to built-in safety guardrails, a gap that reflects differing safety policies across AI labs on whether models should perform offensive security tasks at all.

2. Access to Gemini 3.5 Flash Cyber remains restricted. Google limits the model to a pilot program for governments and trusted partners because the technology possesses dual-use capabilities that could be exploited for malicious purposes.

3. GitHub Copilot administrators face an additional configuration burden. Business and Enterprise users must manually enable the Gemini 3.6 Flash Preview policy before organization members can access the new model.

4. The rollout of Gemini 3.6 Flash within GitHub Copilot is not instantaneous. Google is conducting the integration gradually, which may delay access for some users depending on their specific account status.