Nano Banana 2 & Pro Playground
Experience the next generation of AI image editing and generation powered by Gemini 3 Pro.
Input
Add image from URL or paste from clipboard
Result
Enter a prompt or select an example to start
What would you like to do next?
Your request will cost $0.039 per image. For $1.00, you can run this model 25 times.
What is Nano Banana 2 Pro?
Nano Banana 2 Pro is the next-generation AI image editor powered by the groundbreaking Gemini 3 Pro model.
Why Choose Nano Banana 2 Pro?
3D Object Editing in 2D
Manipulate objects within flat images with full perspective awareness. Rotate, resize, and relocate items while automatically adjusting lighting and shadows to match the scene.
Multi-Image Composition
Upload up to 14 reference images to guide the AI. Perfect for maintaining brand style, character consistency, or combining elements from multiple sources into a cohesive whole.
Context-Aware Editing
The Gemini 3 Pro engine understands the context of your image. It ensures that any additions or changes respect the original style, lighting direction, and physical properties of the scene.
High-Fidelity Text Rendering
Say goodbye to gibberish. Nano Banana 2 Pro renders legible, stylistically accurate text on signs, labels, and documents within your generated images.
Nano Banana 2 Pro vs. The Competition
| Feature | Nano Banana 2 Pro | Nano Banana 1 | Standard AI Editors |
|---|---|---|---|
| AI Model | Gemini 3 Pro | Gemini 2.5 | Various |
| Reference Images | Up to 14 | Single | 1-3 |
| 3D Awareness | Limited | ||
| Text Rendering | High Fidelity | Basic | Inconsistent |
Built for Professionals & Enterprise
E-commerce
Instantly refresh product photography. Change backgrounds, add props, or update models without a reshoot, all while keeping the product 100% authentic.
Marketing
Create brand-consistent ad assets at scale. Use reference images to lock in your brand's color palette and style across hundreds of variations.
Creative Design
Accelerate concept art and storyboarding. Maintain character consistency across different scenes and poses with advanced multi-image referencing.