# Nano Banana 2 Prompt Guide & Library

> The comprehensive technical manual, optical prompt handbook, and curated repository of 2,020 production-ready prompts engineered exclusively for Google's Nano Banana 2 image generation and visual transformation model.

## Overview

Nano Banana 2 is Google's state-of-the-art multimodal generative model designed for high-fidelity visual generation and rapid editing workflows. Key model capabilities include:
- **Advanced World Knowledge**: Accurate understanding of physical materials, geographical environments, architectural eras, and cultural iconography.
- **Subject & Identity Consistency**: Persistent character features, wardrobe retention, and coherent spatial elements across multi-turn generation and transformation turns.
- **Precise Instruction Following**: Faithful interpretation of complex natural language compositions, multi-subject blocking, and strict compositional parameters.
- **Text Rendering & Typography**: Photorealistic rendering of legible in-image typography, signage, labels, headlines, and apparel graphics.
- **Dual Creation Workflows**: Seamless support for both pure text-to-image generation and reference-guided image-to-image editing, restyling, and face preservation.

This platform provides a searchable, open knowledge base and structured prompt dataset containing 2,020 verified prompts categorized across 16 visual domains.

---

## Dataset Statistics & Classification Taxonomy

- **Total Curated Prompts**: 2,020
- **Unique Categories**: 16
- **Verified JSON-Structured Prompts**: 347
- **Reference-Guided Prompts**: 492
- **Direct Text-to-Image Prompts**: 1,528

### Category Breakdown

1. **Portraits & People** (169 prompts): Cinematic portraits, character studies, studio headshots, natural skin textures, and flattering lighting ratios.
2. **Posters & Typography** (169 prompts): Graphic design layouts, Swiss grid typography, editorial book/album covers, and typographic posters.
3. **Product Photography** (167 prompts): Studio packshots, cosmetic bottles, rim-lit pedestals, liquid splash dynamics, and e-commerce hero shots.
4. **Storyboards & Comics** (166 prompts): Multi-panel comic strips, graphic novel narrative sequences, cinematic blocking, and dynamic storytelling frames.
5. **Infographics & Education** (166 prompts): Cutaway diagrams, anatomical visualizers, architectural blueprints, exploded views, and educational schematics.
6. **Fashion & Beauty** (166 prompts): Haute couture lookbooks, runway motion, editorial cosmetics, textile draping, and designer collections.
7. **Social Media & Editorial** (164 prompts): Authentic candid lifestyle snapshots, smartphone optics, selfie angles, influencer aesthetics, and everyday narratives.
8. **Image Editing & Transformation** (162 prompts): Image-to-image instructions, outfit swaps, background replacements, face-preservation workflows, and restyling tasks.
9. **Game Assets & Fantasy** (160 prompts): Isometric game sprites, RPG character concept sheets, creature designs, fantasy environments, and UI icon sets.
10. **Food & Beverage** (153 prompts): Commercial culinary photography, steam, condensation, plated dishes, cocktail garnishes, and tabletop styling.
11. **Architecture & Interiors** (136 prompts): Modernist, brutalist, and Scandinavian spaces, architectural facades, residential lofts, and spatial lighting.
12. **Concept Art & Sci-Fi** (121 prompts): Cyberpunk cityscapes, retro-futurism, spacecraft, mecha, planetary landscapes, and cinematic environment concept art.
13. **Cinematic Photography** (47 prompts): 35mm anamorphic film frames, widescreen aspect ratio stills, shallow focus, and dramatic color grading.
14. **Web & UI Design** (38 prompts): Modern SaaS landing pages, mobile app interfaces, executive dashboard mockups, and responsive web layouts.
15. **Advertising & Commercial** (24 prompts): High-impact marketing key visuals, billboard campaigns, and branded advertising imagery.
16. **Automotive & Vehicles** (12 prompts): Sports cars, hypercars, vintage classics, dynamic tracking shots, and automotive studio lighting.

---

## Nano Banana 2 Prompting Principles

### 1. Concrete Physical Specifications
Describe tangible surfaces, materials, and textures rather than vague superlatives:
- *Recommended*: "Matte brushed aluminum body with knurled steel dials, micro-scratches on glass bezel, soft diffused daylight."
- *Avoid*: "Hyper-realistic 8k masterpiece supreme ultra-detailed."

### 2. Optical & Camera Direction
Anchor scene perspective and optical characteristics explicitly:
- **Focal Lengths**: 24mm (wide environmental view), 35mm (documentary/street), 50mm (natural human eye), 85mm (compression portrait), 100mm+ macro (product details).
- **Aperture & Depth of Field**: f/1.4-f/2.8 for shallow depth of field and soft background bokeh; f/8-f/11 for deep architectural sharpness.
- **Angle & Framing**: Low-angle heroic perspective, bird's-eye overhead flat lay, eye-level candid, or Dutch angle tension.

### 3. Lighting Architecture
Direct light sources and quality to define mood and volume:
- **Key & Fill Ratio**: High-contrast dramatic chiaroscuro (4:1 or 8:1) vs. clean, soft commercial fill (2:1).
- **Color Temperature**: Warm golden hour (3200K tungsten) paired with cool blue ambient shadow fill (6500K).
- **Accents**: Subtle rim light / edge separation along shoulders, silhouettes, and glass rims.

### 4. Reference-Guided Transformations
When transforming existing imagery:
- Use consistent identifier notation: `Reference Image 1`, `Reference Image 2`.
- Explicitly declare what to preserve and what to alter (e.g., "Preserve facial features and identity from Reference Image 1, but replace the background with a minimalist concrete art gallery").

---

## Prompt Data Schema

Each prompt in the repository conforms to the following structured JSON schema:

```typescript
interface PromptItem {
  id: string;                      // Unique identifier, e.g. 'nb2-0001'
  title: string;                   // Concise title
  prompt: string;                  // Full prompt string or serialized JSON prompt
  category: string;                // Primary semantic category (one of 16)
  subcategory?: string;            // Secondary topic classification
  description?: string;            // Technical description or execution note
  requiresReference: boolean;      // True if the prompt requires a reference image
  techniques: string[];            // Authentic techniques used (e.g. '85mm Portrait Prime')
  tags: string[];                  // Descriptive keywords and metadata
  breakdown: {                     // Five-factor technical breakdown
    subject: string;               // Hero subject and action
    camera: string;                // Lens, sensor, angle, and framing
    lighting: string;              // Light setup, direction, and color temperature
    composition: string;           // Geometric rules, framing, and depth layers
    mood: string;                  // Atmosphere, emotion, and aesthetic tone
  };
}
```

---

## Machine-Readable Resources & Endpoints

- **Interactive Web Guide & Search UI**: https://banana2-prompts.ivvideo.com/
- **Full Dataset Endpoint (JSON, ~5MB)**: https://banana2-prompts.ivvideo.com/data/prompts.json
- **Model & Category Statistics (JSON)**: https://banana2-prompts.ivvideo.com/data/stats.json
- **XML Sitemap**: https://banana2-prompts.ivvideo.com/sitemap.xml
- **Robots Policy**: https://banana2-prompts.ivvideo.com/robots.txt