Google launches cost-efficient Gemini models to scale AI agents and secure code
The tech giant introduces lower-priced models with improved token efficiency and speed, positioning them against rivals in the agentic AI and security markets.
Google has released three new artificial intelligence models designed to enhance efficiency, latency, and reliability for building AI agents at scale. The new suite includes Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber, marking a strategic push to lower costs for developers while improving performance in coding and knowledge-based tasks.
Gemini 3.6 Flash builds on developer feedback from its predecessor, delivering improved capabilities in coding and knowledge work while consuming 17% fewer output tokens according to the Artificial Analysis Index. The model is priced at $1.50 per million input tokens and $7.50 per million output tokens, representing a cost reduction compared to Gemini 3.5 Flash. This enhanced efficiency is combined with stronger Frontier Safety safeguards against Chemical, Biological, Radiological, and Nuclear misuse, as well as cyber offense attempts.
For high-throughput workflows, Google introduced Gemini 3.5 Flash-Lite, positioned as the fastest model in the 3.5 series. Running at 350 output tokens per second, it is priced at $0.30 per million input tokens and $2.50 per million output tokens. The model features built-in computer use capabilities and configurable thinking levels, allowing developers to balance speed and cost. It outperforms Gemini 3 Flash on specific benchmarks, including SWE-Bench Pro and OSWorld-Verified.
A distinct release is Gemini 3.5 Flash Cyber, which is being made available exclusively to governments and trusted partners via a limited-access pilot program through CodeMender. Fine-tuned to detect and fix cybersecurity vulnerabilities, the model aims to provide a cost-effective alternative to larger, compute-heavy solutions in the security sector. This release occurs amid a broader industry race to develop affordable AI security tools, with competitors such as Microsoft and China’s Z.ai also advancing their own capabilities.
Beyond these releases, Google confirmed that Gemini 3.5 Pro is currently in testing with partners, with broad availability planned for when ready. The company also noted that pre-training has begun for Gemini 4, described as the team’s most ambitious run yet, signalling continued investment in next-generation model architecture.
