google-gemini/gemini-image-editing-nextjs-quickstart ? reverse-engineered prompt

Build me a simple web app where people can create and edit images with Gemini 2.0 Flash. I want a clean, responsive interface where someone can type a prompt to generate a new image, or upload an image and tell it what changes they want in plain language. It should keep the conversation going so users can refine the same image step by step without starting over every time.

Please make it easy to switch between creating and editing, show the result clearly, and let users download the final image. Use a modern Next.js frontend with a polished, friendly layout, and make sure it works well on desktop and mobile. I’ll add my Gemini API key in an environment file, so wire it up that way. If you need to check the latest Gemini docs online while building it, go ahead.