Nano Banana AI Image Generator is the Uni-1 image workflow with the Nano Banana model option preselected. You can generate from text, upload an image for editing, choose the aspect ratio, and keep recent jobs in the same workspace.
Google describes Nano Banana as Gemini native image generation capabilities that can work conversationally with text, images, or both. That makes it useful for text-to-image, image-to-image edits, multi-step refinement, and prompts that try to preserve a subject while changing the surrounding scene.
The exact provider behavior can depend on the runtime API and the selected Nano Banana family model, so this page avoids promising unsupported controls. The current workflow stays focused on prompt, upload, ratio, preview, download, and practical prompt patterns for local editing, style transfer, and consistent subjects.