Advanced Features Guide

Advanced Capabilities Overview

Nano Banana Pro 2 offers advanced capabilities beyond basic text-to-image generation. This guide covers the features that make it suitable for professional creative workflows.

---

Thinking-Powered Composition

Nano Banana Pro 2 uses an intelligent reasoning engine that plans composition before creating the image. This "thinking" process:

1. Analyzes your prompt and extracts key requirements
2. Designs multiple layout options
3. Selects optimal composition based on the request
4. Generates the final image with precise element placement

The thinking process is automatic and cannot be disabled. For faster results with simpler scenes, consider using standard prompts rather than highly complex multi-element requests.

When Thinking Mode Excels

• Complex scenes with multiple subjects
• Scenes requiring specific spatial relationships
• Compositions with specific focal points
• Multi-panel sequential art (comics, storyboards)

Example Prompts

Complex Scene:

A bustling Tokyo street market at night with neon signs reflecting on wet pavement, a ramen shop in the foreground with warm light spilling onto the street, a salaryman walking past checking his phone, cherry blossoms floating in the air. Wide-angle shot from street level, cinematic lighting.

Sequential Art:

Two-page full-color manga comic (12 panels): Doraemon + Nobita battle Godzilla in coastal Japan. Crisp lines, cinematic pacing, short US-English bubbles/SFX, no gore. Gadget accident starts conflict; teamwork/gadgets escalate.

---

Multi-Reference Image Composition

Nano Banana Pro 2 supports up to 14 reference images for complex consistency editing. This enables sophisticated workflows that were previously difficult or impossible with AI image generation.

Reference Image Types

Object Images: Reference images of specific objects to be included in the generated scene

• Products for e-commerce photography
• Real-world objects to place in new contexts
• Specific items that must appear in the output

Character Images: Reference images of people or characters for consistency

• Model references for fashion photography
• Character design references for comics
• Brand mascot references for marketing

Model-Specific Limits

Example Workflows

Product Photography Composition

1. Upload product image (e.g., a dress on white background)
2. Upload model image (e.g., a person standing)
3. Use prompt: "Create a professional e-commerce fashion photo. Let the woman wear this blue floral dress. Generate a realistic, full-body shot with the lighting adjusted to match the outdoor environment."

Brand Mascot Development

1. Upload multiple angles of a character or mascot design
2. Upload reference images for desired settings
3. Generate consistent compositions in various scenarios

Multi-Character Scene

1. Upload individual character reference images
2. Describe the scene with all characters
3. Maintain identity consistency across the scene

---

Real-Time Web Knowledge Grounding

Enable the Web Search toggle to generate images with current, accurate information. This feature connects the model to live web data for:

Use Cases

• Weather visualizations: Current forecast data displayed graphically
• Sports graphics: Live match scores, team standings, player stats
• News visualization: Current events rendered as images
• Location accuracy: Real landmarks and geographical features
• Product accuracy: Current product specifications and appearances

Example Prompts

Weather Forecast:

Visualize the current weather forecast for the next 5 days in San Francisco as a clean, modern weather chart. Add a visual on what I should wear each day.

Sports Score:

Make a simple but stylish graphic of last night's Champions League match featuring Arsenal.

Real-World Location:

Create a photorealistic image of the Golden Gate Bridge in San Francisco under current weather conditions.

Attribution Requirements

When using web search grounding, you may need to provide attribution to sources used for generated content. Always verify that generated content accurately represents real-world information.

---

Image-to-Image Editing

Beyond simple variations, Nano Banana Pro 2 supports sophisticated editing through image upload.

Adding Elements

Upload an image and describe what to add:

Using the provided image of my cat, please add a small, knitted wizard hat on its head. Make it look like it's sitting comfortably and not falling off.

Selective Changes (Inpainting)

Change specific elements while preserving others:

Using the provided image of a living room, change only the blue sofa to be a vintage, brown leather chesterfield sofa. Keep the rest of the room, including the pillows on the sofa and the lighting, unchanged.

Style Transfer

Transform images to different artistic styles:

Transform the provided photograph into the artistic style of Vincent van Gogh's "Starry Night". Preserve the original composition but render with swirling, impasto brushstrokes and deep blues and bright yellows.

High-Fidelity Detail Preservation

When editing, describe critical details that must remain unchanged:

Take the first image of the woman. Add the logo onto her black t-shirt. Ensure the woman's face and features remain completely unchanged. The logo should look naturally printed on the fabric, following the folds of the shirt.

---

Sketch to Masterpiece

Transform rough hand-drawn concepts into polished visuals. Upload any sketch or doodle and describe your desired refinement.

Ideal For

• Industrial design: Concept sketches to product renders
• Architecture: Hand-drawn plans to photorealistic renders
• Product design: Rough mockups to professional visuals
• Illustration: Sketches to finished artwork

Example

Input: Rough pencil sketch of a sports car Prompt: "Turn this rough pencil sketch into a polished photo of the finished concept car in a showroom. Keep the sleek lines and low profile from the sketch but add metallic blue paint and neon rim lighting."

Best Results

• Use clean, clear line drawings
• Include key structural elements in the sketch
• Be specific about desired materials and lighting in prompt
• Reference desired style (photorealistic, stylized, etc.)

---

Character Consistency

Maintain consistent character appearances across multi-panel stories or variations.

Workflow

1. Upload a high-quality reference with clear facial features
2. Request specific angles one at a time
3. Reference previous generations in subsequent prompts for consistency

Example: 360° Character Views

Initial:

A studio portrait of this person, front view, against white background.

Follow-up:

Same character as the provided reference image, in profile looking right, studio lighting.

Continue:

Same character, three-quarter view, slightly elevated angle.

Tips for Best Consistency

• Use high-quality reference images with clear, front-facing features
• Keep backgrounds simple in reference images
• Request one view change at a time
• Use consistent lighting descriptions across generations
• Reference the character description from previous outputs

---

Resolution & Quality Options

Resolution Tiers

Aspect Ratio Guide

---

Prompt Engineering Tips

For Realistic Photography

Use photography terminology to guide lighting, composition, and mood:

A photorealistic close-up portrait of an elderly ceramicist. The scene is illuminated by soft, golden hour light streaming through a window. Captured with an 85mm portrait lens, resulting in a soft, blurred background (bokeh).

For Stylized Work

Be explicit about artistic style and background:

A kawaii-style sticker of a happy panda. Bold, clean outlines, simple cel-shading, vibrant color palette. The background must be white.

For Product Shots

Describe lighting setup and camera angle:

High-resolution product photograph on a polished concrete surface. Three-point softbox setup for soft, diffused highlights. Slightly elevated 45-degree shot to showcase clean lines.

For Complex Scenes

Break down the prompt into components:

1. Subject and action
2. Setting and environment
3. Lighting and mood
4. Camera angle and technical specs

Iteration Strategy

1. Start with a broad, descriptive prompt
2. Generate and evaluate the result
3. Make targeted adjustments in follow-up prompts
4. Iterate until satisfied

---

Related Guides

• Creating Your First Image — Step-by-step tutorial
• Prompt Structure — How to structure effective prompts
• Character Reference — Using character references
• Style Reference — Applying artistic styles