Gemini 3.5 Flash-Lite is Google DeepMind's lowest-cost model in the current lineup, positioned below 3.6 Flash for high-volume, latency-sensitive workloads. At $0.30 input / $2.50 output per million tokens, it targets agentic search, document processing, and classification tasks where throughput matters more than peak intelligence. It outputs at roughly 350 tokens per second and shares the 1M-token context window of its Flash siblings.