An overview of the features and pricing of “Gemini 3.1 Flash-Lite,” the fast and lightweight AI model announced by Google on March 3, 2026. This article concisely summarizes facts such as the overwhelmingly low cost of $0.25 per 1 million input tokens, response speeds 2.5 times faster than the previous generation, and how to utilize “Thinking levels” to adjust reasoning depth.