nano-banana-pro

To generate images from text prompts using the Gemini 3 Pro API, provide a descriptive prompt and specify your desired resolution. You can configure the output to 1K, 2K, or 4K, and the generated image is saved directly to your current directory.

Yes, you can edit an existing image file with Gemini 3 Pro by passing the file directly via the --input-image parameter. You should not read the image file first; simply provide the file path along with your specific editing instructions to modify aspects like background or style.

Yes, you need an API key to use the Gemini 3 Pro image generation API. You must provide it either by passing the --api-key flag in your command or by setting the GEMINI_API_KEY environment variable in your operating system.

Text-to-image generation supports configurable resolutions of 1K, 2K, or 4K for high-resolution outputs. The default resolution is set to 1K, but you can explicitly specify higher resolutions like 4K for detailed portfolio images or branding assets.

The best way to modify the background of an existing image is to use image-to-image editing with Gemini 3 Pro. Provide the original file through the --input-image parameter and apply specific text instructions to replace or alter the background directly without reading the file beforehand.