Google has recently rolled out a rapid, economical AI image generation and editing model tailored for developers: Nano Banana 2 Lite (formally known as Gemini 3.1 Flash-Lite Image). This model prioritizes speed and cost-efficiency, capable of producing 1K resolution images in roughly 4 seconds, with the average cost per thousand images via API calls being just $0.034. It is crafted for high-volume commercial uses, swift prototyping, and automated asset creation processes. Nano Banana 2 Lite can generate scene sketches grounded in global knowledge, preliminary data visualization drafts, and location-specific simulations. It also boasts character consistency preservation, ensuring text within generated images is rendered clearly. Nevertheless, the model presently only supports 1K resolution output, lacking 2K or 4K alternatives, and has constraints in rendering exceptionally small text and the precision of infographic data. In the realm of text-to-image conversion, Nano Banana 2 Lite boasts an Elo rating of 1251, outperforming the original Nano Banana (1151) and edging out Nano Banana Pro (1245). At present, the model is officially accessible on Google AI Studio, Gemini API, and the Gemini Enterprise Agent Platform.