Google DeepMind introduces Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, two new models delivering enhanced reasoning and coding performance at the same speed and cost as Gemini 3.7 Flash. Raluca Ada Popa, Gemini Security Lead at Google DeepMind, notes this marks the third Flash release in six weeks. Gemini 3.8 Flash serves as an intelligent workhorse for software engineering, agentic tasks, and multi-step reasoning, priced at $0.75 per million input tokens and $3.75 per million output tokens.

It outperforms larger frontier models on benchmarks like Deep SWE v1.1, Vals Finance Agent V2, and Harvey's Legal Agent Benchmark, achieving 54.9% on HLE-Verified. Gemini 3.8 Flash Cyber, available via the Fairwind Program to trusted defenders, excels in vulnerability detection and automated patching, surpassing prior models and larger frontier models on Cyber Gym and internal benchmarks across 20 programming languages, with a success rate exceeding 70%. Both models share foundational intelligence and are accelerated by long-running agentic loops for recursive refinement.