Google Unveils Three Gemini AI Models to Expand AI and Cybersecurity Push

Google has introduced three new Gemini artificial intelligence models, strengthening its push to compete in the rapidly evolving AI market. The latest releases focus on speed, affordability, and specialized performance across software development, automation, and cybersecurity.

Besides improving efficiency, the company aims to lower operating costs for developers while expanding AI capabilities for enterprise users. The rollout also highlights Google’s growing focus on defensive cybersecurity as competition with OpenAI, Anthropic, Meta, and emerging AI firms continues to intensify.

New Models Target Performance and Lower Costs

Gemini 3.6 Flash leads the new lineup with stronger performance across coding, finance, and complex reasoning tasks. Additionally, Google introduced Gemini 3.5 Flash-Lite for lightweight workloads, including AI agents that automate repetitive digital tasks.

Consequently, developers can complete similar workloads while using fewer computing tokens, reducing overall expenses. Google stated the latest Flash models require roughly 17% fewer tokens than earlier versions, improving efficiency without sacrificing quality.

Cybersecurity Takes Center Stage

Google also unveiled Gemini 3.5 Flash Cyber, a security-focused model designed to identify and repair software vulnerabilities. Moreover, internal testing found previously undetected weaknesses inside the V8 JavaScript Engine, demonstrating stronger defensive capabilities.

However, Google will initially restrict access to governments and trusted partners to reduce misuse risks. Meanwhile, the company continues testing Gemini 3.5 Pro, its flagship model, before a broader public release.

Related: US Capital Research Raises Metaplanet Voting Stake to 10.63%