What is Nano Banana?

What “Nano Banana” means here

Nano Banana Pro 2 is the next-generation image model bundled with this app: built for fast turnaround, strong instruction following, and high-resolution output (including native 4K-class results depending on your plan and settings).

It is not a single “magic button”—it works best when you give clear intent, camera and lighting language, and constraints (what to avoid, what must stay consistent).

What it does well

• Photoreal scenes when you specify lens, lighting, and materials.
• Typography and UI mockups when you describe text, placement, and style explicitly.
• Edits and variations when you keep the main subject stable and change one axis at a time.
• Multi-image workflows (style or character reference) when references are high quality and on-brief.

What to watch for

• Hands, teeth, and fine patterns may need a second pass or a tighter prompt.
• Exact trademarked characters or logos can be restricted—describe the look instead of the brand.
• Legibility of long text improves when you shorten copy, increase contrast, and specify font style.

How this doc set is organized

• Prompts — how to write and structure language the model understands.
• Image tools — references, style, film looks, and lighting vocabulary.
• Parameters — aspect ratio, quality, stylization, chaos, seed, and “raw” looks.
• Using the website — gallery, search, subscriptions, and account tools.

When you are ready, open the Getting started guide and run your first session end to end.

Key Capabilities

Text-to-Image Generation

Describe the scene rather than just listing keywords. The model understands nuanced language and produces better results with descriptive, narrative paragraphs.

Best practices:

• Use photography and cinematography language
• Specify lighting type, direction, and mood
• Describe camera angle, lens type, and depth of field
• Include material textures and surface details

Image-to-Image Editing

Upload a reference image and describe your goal in plain language. The model maintains consistency with the original image's style, lighting, and perspective.

Common tasks:

• Adding or removing elements
• Style transfer and artistic rendering
• Selective inpainting (change specific areas)
• Multi-image composition for product mockups

Multi-Reference Composition

Support for up to 14 reference images enables:

• Product-on-model photography
• Character consistency across scenes
• Brand mascot development
• Complex visual storytelling

Real-Time Web Knowledge

Generate images with live, accurate information:

• Current weather, sports, and news
• Real locations and landmarks
• Verified facts and trending topics

Sketch to Masterpiece

Transform rough hand-drawn sketches into polished visuals:

• Industrial design concepts
• Architectural rendering
• Product prototyping
• Illustration refinement

Prompt Writing Best Practices

1. Be descriptive, not just listing — "A cozy coffee shop with warm afternoon light streaming through windows" works better than "coffee shop, warm light"

2. Use photography language — Specify shot type, lens, lighting, and mood

3. Describe what you want to AVOID — Negative prompts help exclude unwanted elements

4. Iterate and refine — Start broad, then add specific details in follow-up prompts

5. For complex scenes, break it down — "First create a forest background, then add a stone altar, then place the sword"