Google's fastest frontier model, announced at I/O 2026. Gemini 3.5 Flash uses dynamic thinking by default, automatically deciding when to reason deeply versus respond quickly. It surpasses the larger 3.1 Pro on coding, agentic, and multimodal benchmarks while delivering roughly 4x faster output. The 1M token context window handles large codebases and document sets comfortably. Positioned as a high-capability model at Flash-tier pricing, it hits a strong performance-per-dollar ratio for production workloads.