nano-banana-pro

To generate images from text prompts using Gemini 3 Pro, provide a natural language prompt to the Nano Banana Pro script via the uv CLI runner, which outputs a PNG file. You must configure the GEMINI_API_KEY environment variable beforehand.

Yes, you can edit multiple existing images with Gemini 3 Pro by supplying up to 14 input images alongside a prompt to the Nano Banana Pro script. The script composes the results and outputs the edited PNG files.

Running image generation with the uv CLI requires setting the GEMINI_API_KEY environment variable and installing the google-genai and pillow dependencies. The script uses the uv runner workflow to execute the generation process.

The best way to create concept art and marketing visuals quickly is using Gemini 3 Pro Image generation to produce rapid visual outputs from natural language prompts. This approach supports design workflows requiring fast visual exploration.

When editing images, the script accepts up to 14 input images for composition and outputs the resulting files strictly as PNG format. This ensures compatibility for design and concept art workflows.

Running the script fails because Gemini 3 Pro image generation requires authentication via the GEMINI_API_KEY. You must provide this key either through an environment variable or by passing the --api-key argument during execution.