Nano Banana is a Google Gemini–powered AI image editing tool that can transform photos and maps into stylized 3D or illustrated renderings. It supports multi-turn prompts and editing, making adjustments like expressions, clothing, backgrounds, and even combining multiple characters. Below is a concise guide to get started and common workflows.

Quick start

• Access a Gemini-based interface (Gemini or compatible AI Studio/chat).
• Upload one or more images or start from a text prompt to generate images in your preferred style.
• Provide targeted prompts to shape the result, then refine via follow-up prompts to adjust details like pose, expression, outfit, or background.

Typical workflows

• Single-image edit: Upload a photo, describe the desired changes (e.g., “I want the background changed to a cityscape at night; keep facial expression neutral.”), then iterate until satisfied.
• Multi-turn refinement: After an initial render, add more specific adjustments (e.g., “make the hair longer and brighter,” “place the character on a computer desk with no text visible”).
• Composite scenes: Describe multiple characters or elements and request a cohesive scene; refine as needed to balance lighting and composition.
• Output styles: Realistic, painterly, game-art, or other stylistic cues. You can specify lighting, color tone, and environment to guide the render.

Practical tips

• Be explicit about the target style and environment to improve fidelity.
• Use concise, incremental prompts to reduce misunderstandings; leverage context from previous prompts to maintain continuity.
• If a result deviates, reframe the instruction (e.g., clarifying which element needs to change or which asset should be foreground).

If you’d like, share the image(s) you want to edit and the exact changes you’re aiming for, and a tailored, step-by-step prompt sequence can be crafted.