Result data is only kept for 3 days

Guide
Create next-generation AI images with Nano Banana 2. Produce detailed visuals, style variations, and high-resolution assets up to 4K.
Describe the image you want to generate in detail, including style, colors, composition, etc. The more detailed, the better the result.
You can upload a reference image or select from examples. AI will generate new images based on both the reference and your description.
Click "Generate Image" button, and the system will create high-quality images using Google Gemini AI.
Create product scenes, campaign directions, and branded visual variations from a focused brief.
Use clear references and instructions to explore new scenes while preserving important visual traits.
Produce illustrations, thumbnails, and fast visual concepts sized for different publishing needs.
Nano Banana 2 is the newer generation of Google's Gemini image model, with better prompt understanding and detail rendering than Nano Banana Pro. It generates at 1K, 2K or 4K and takes up to 4 reference images.
Nano Banana 2 is the default choice for new work. Nano Banana Pro remains useful when you are extending an existing set and want to stay consistent with images you already generated on it.
Upload the same reference images every run and describe only what changes β pose, outfit, setting. Character consistency comes from stable references, not from repeating the description of the face in words.
When the image will be printed, cropped hard, or used as a hero visual. For social posts and thumbnails, 1K or 2K is indistinguishable at display size and costs meaningfully less per image.