Google new version of Nano Banana online

On November 20th, Google's new multi-modular AI model, the second-generation Nano Banana (coded Gemini 3.0 Pro Image) update, and the new official designator is Gemini 3 Pro Image. Unlike the traditional image model, Gemini 3 Pro Image is a model of reasoning that is internal to the image before it is generated, as if it was thinking about how best to do the job. This has resulted in significantly improved image quality, greater accuracy and better multilingual long text rendering. According to official bloggers, Gemini 3 Pro Image was designed to address the most challenging image generation tasks. The model is now on-line on the Vertex AI platform and can also be used in Google AI SDK (which provides a single Python and Go interface)。