Google has introduced three new models in its Gemini series: Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber, designed to improve token efficiency, latency, and reliability for production AI agents.
- Gemini 3.6 Flash reduces output token usage by 17% compared to 3.5 Flash on the Artificial Analysis Index, with up to 65% reduction in DeepSWE benchmarks, while lowering costs to $1.50/1M input and $7.50/1M output tokens.
- Gemini 3.5 Flash-Lite delivers 350 output tokens per second according to the Artificial Analysis Index, offering a price of $0.3/1M input and $2.5/1M output tokens for high-throughput agentic workflows.
- Gemini 3.5 Flash Cyber is a specialized model paired with CodeMender for cybersecurity applications, currently available in a limited-access pilot program for governments and trusted partners.
These releases aim to provide developers with more cost-effective and efficient tools for scaling complex agentic tasks, document processing, and security vulnerability detection.