How to Use Nano Banana: Complete Guide 2025
Learn how to use Google's revolutionary Nano Banana (Gemini 2.5 Flash Image) with our step-by-step tutorial collection. Master AI image generation, editing, video creation, and developer integration from beginner to advanced level.
The Technology Behind Nano Banana
In this Nano Banana Guide, we break down how Nano Banana works under the hood: a Google-built AI image engine that fuses prompts and reference signals to produce consistent visuals. Understanding the prompt-to-pixel pipeline helps you craft better instructions and achieve stable, repeatable outcomes.
Understanding the AI Model
Powered by Gemini 2.5 Flash Image, Nano Banana aligns text and image embeddings with attention-based consistency. It preserves identity across edits while allowing style, lighting, and background changes—ideal for character consistency, product shots, and multi-image storytelling.
Why This Guide Helps You Master Nano Banana
Knowing how Nano Banana works lets you structure prompts, choose reference images, and iterate effectively. This Nano Banana Guide provides an authoritative foundation so you can learn fast and generate dependable results.
Developer Tutorial by Google DeepMind
Official tutorial from Patrick at Google DeepMind covering API integration, Python/JavaScript examples, and best practices for developers.
Easy Tutorial for Beginners
Step-by-step guide covering basic image generation, reference images, product placement, and adding clothes to models.
Video Generation Tutorial
Complete guide to creating videos with Nano Banana, including prompt formulas, Google Flow integration, and frame-to-video techniques.
Photoshop Alternative Guide
Advanced tutorial showing how Nano Banana can replace Photoshop for image manipulation, combining multiple images, and complex editing tasks.
Developer Tutorial by Google DeepMind
Complete developer guide from Patrick at Google DeepMind
💡 What You'll Learn in This Tutorial
- How to set up Google AI Studio for Nano Banana development
- API integration with Python and JavaScript examples
- Image generation and editing techniques
- Best practices for production applications
Getting Started with AI Studio
The best place to get started with Nano Banana as a developer is in Google AI Studio. This is our playground, where you can experiment with all our models and then also move this into your apps later.
Quick Setup Steps:
- Go to
ai.studioand sign up with your Google account - Select Nano Banana in the model picker
- Create images or upload images and edit them
- Experiment with prompts like "restore and colorize the image"
API Setup and Integration
To build with Nano Banana, we use the Gemini Developer API and one of the official Google Gen AI SDKs. Here's how to set it up:
1. Create API Key
- In AI Studio, click on "Get API Key"
- Click "Create API Key" (creates Google Cloud project first time)
- Copy and store the key securely
2. Set Up Environment
GEMINI_API_KEY=your_api_key_here3. Install SDK
pip install -U google-genai pillow python-dotenvBasic Image Generation Code
import google.generativeai as genai
from PIL import Image
from io import BytesIO
# Set up client
client = genai.Client()
# Create prompt
prompt = "Create a photorealistic image of an orange cat with green eyes sitting on a couch"
# Generate image
response = client.models.generate_content(
model="gemini-2.5-image-preview",
contents=[prompt]
)
# Save image
for part in response.candidates[0].content.parts:
if hasattr(part, 'inline_data'):
image_data = BytesIO(part.inline_data.data)
image = Image.open(image_data)
image.save("generated_cat.png")Image Editing with Multiple Inputs
You can easily edit images by providing both a prompt and an image:
# Load existing image
image = Image.open("cat.png")
# Edit with prompt
prompt = "Create a street level view of the cat walking in New York City"
response = client.models.generate_content(
model="gemini-2.5-image-preview",
contents=[prompt, image]
)Conversational Image Editing
For iterative editing, use chat sessions:
# Create chat session
chat = client.chats.create(model="gemini-2.5-image-preview")
# First edit
response1 = chat.send_message(["Change the cat to a bengal cat", image])
# Continue editing
response2 = chat.send_message("The cat should wear a funny party hat")Best Practices for Developers
- Be specific: Provide context and intent
- Iterate and refine: Use step-by-step instructions
- Use positive framing: Instead of "no cars", say "empty, deserted street"
- Control camera: Use terms like "wide-angle shot", "macro shot"
- Photo restoration: Great for colorizing old photos with context like "photograph from 1932"
💡 Pro Tip
Nano Banana costs around $0.04 per image. For $1.00, you can create roughly 25 images. Check the official pricing page for up-to-date information.
Easy Tutorial for Beginners
Step-by-step guide for photographers and content creators
🎯 What You'll Master in This Guide
- Free access to Nano Banana through Google Gemini
- Reference image techniques for professional results
- Product placement and fashion photography
- Conversational editing for iterative improvements
How to Access Nano Banana
Google has added Nano Banana into Google Gemini. As long as you have the latest version, you can access it there. It's free, but you do get limited uploads and generations. For more generations, upgrade to the pro account.
Access Steps:
- Go to Google Gemini
- Choose the Flash 2.5 model
- You now have access to the image model
1. Reference Image Technique
This technique is super powerful and gives you great control over AI backgrounds. Here's how to use it:
Step-by-Step Process:
- Upload Background: Click "upload a file or image" and choose a background you like
- Add Product: Upload an image of the item you want to add (like a watch)
- Write Prompt: Describe how you want the final image to look
- Generate: Wait about 20 seconds for results
Example Prompt:
"Create a dramatic image of this watch with lightning in the background and rocks. Make it super moody with dust coming from the bottom of the watch, hovering in the air."
2. Conversational Editing
You can refine your images by continuing the conversation:
- "Make the watch hover higher in the air"
- "Keep the watch the same size but move it up"
- "Adjust the lighting to be more dramatic"
3. Product in Hands Technique
Nano Banana excels at placing products in people's hands naturally:
Best Practices:
- Clear Labels: Use products with prominent, clear text for better label reproduction
- Environment Context: Specify the setting ("rainforest environment", "urban setting")
- Camera Details: Add camera specifications for professional results
Example Prompt:
"Show this bottle being held by a hand coming out of the ground, with dirt and grass around it in a rainforest environment. Shot with a Canon EOS R5."
4. Adding Clothes to Models
One of the most impressive features - easily add clothing to people or AI avatars:
Simple Process:
- Upload an image of clothing
- Upload an image of a person
- Use simple prompts like "Put this t-shirt on the man"
- Continue adding items: "Replace his shorts with these jeans"
Key Success Tips for Beginners
- Image Quality: Use high-resolution, clear images for best results
- Specific Prompts: Be detailed about what you want
- Lighting Consistency: Nano Banana automatically matches lighting and shadows
- Text Clarity: For products with text, ensure labels are clear and prominent
- Iteration: Don't hesitate to refine with follow-up prompts
Speed Advantage
Nano Banana is significantly faster than ChatGPT-4 for image generation - typically 20 seconds vs 1+ minutes.
Video Generation Tutorial
Learn to create cinematic videos with Nano Banana
Accessing the Video Model
To access video creation in Nano Banana:
- Go to Google Gemini Flash 2.5
- Click on the "Tools" section
- Select "Create with VA" (above Nano Banana option)
The Perfect Video Prompt Formula
For cinematic results, follow this proven structure:
5-Part Prompt Structure:
- Main Shot & Camera Movement: "Wide-angle dolly shot"
- Subject & Action: "Warrior in black fur coat turning to face"
- Setting & Environment: "Snowy battlefield"
- Visual Style: "Ultra detailed, cinematic"
- Audio: "Deep rumble of dragon breathing, wind howling"
Example Prompts:
Epic Scene:
"Wide-angle dolly shot of a warrior in a black fur coat turning to face a massive ice dragon on a snowy battlefield. Ultra detailed with deep rumble of dragon breathing and wind howling."Cute Scene:
"Close-up shot of a puppy splashing in a drink of milk tea, with the drink flowing out and the dog licking the milk tea on a wooden table. High quality, light and bright, cute. Audio: splashing milk tea sounds."Overcoming Generation Limits
The main limitation: Gemini only gives 3 daily video generations. Here's how to get more:
Use Google Flow
- Access: Included in your Google One subscription
- Credits: 1,000 credits (each video = 20 credits = ~50 videos)
- Models: Access to V2 and V3 models
- Cost Saving: Use V2 fast version to keep credits at 20 per generation
Frame-to-Video Technique
Advanced technique using Google Flow:
Process:
- Create Start Frame: Generate first image in Nano Banana
- Create End Frame: Generate or choose ending image
- Upload to Google Flow: Choose "Frame to Video" option
- Crop Images: Resize to 16:9 aspect ratio
- Add Prompt: Use the 5-part formula
- Select Model: V2 fast for frame-to-video (V3 not available)
⚠️ Frame-to-Video Limitations
- Must use V2 model (not V3)
- No audio generation available
- Requires manual frame creation
Alternative Platforms
For more advanced video features:
- Higfield: Access to Nano Banana, V3, and Cling 2.1
- Cling 2.1: Superior frame-to-video capabilities
- Multiple Models: Best image and video models in one platform
Video Creation Best Practices
- Always use the 5-part prompt formula for best results
- Be specific about camera movements (dolly, pan, zoom)
- Include audio descriptions for immersive results
- Start simple then add complexity
- Plan your scenes before generating
🎬 Pro Tip
Videos generate much faster than ChatGPT alternatives, and the quality is consistently high when using the proper prompt structure.
Photoshop Alternative Guide
Advanced image manipulation that rivals traditional software
Why Nano Banana Can Replace Photoshop
Recent headlines claim Google's Nano Banana is making Photoshop obsolete. Here's why this might be true for many use cases:
- Natural Language Control: No need to learn complex tools
- Intelligent Image Combination: Preserves characteristics automatically
- Realistic Lighting & Shadows: Automatic matching
- Speed: Much faster than manual Photoshop work
1. Combining Two Images
Replace objects between images while preserving quality:
Example Process:
- Upload Images: Person holding a peach + Bose earbuds
- Simple Prompt: "Replace the peach I'm holding with the Bose earbuds case"
- Result: Natural-looking combination with preserved facial features and proper lighting
Key Advantage
Unlike traditional AI, Nano Banana preserves image characteristics instead of recreating everything. Your face stays your face, logos remain readable.
2. Image Reorientation
Create new perspectives of objects that would be impossible in Photoshop:
Technique:
- Upload Product Image: Front view of a backpack
- Request New Angle: "Create a side view of this backpack on a mountain"
- AI Imagination: Nano Banana understands 3D structure and creates realistic side views
Success Factors:
- Simple Objects: Work better than complex organic shapes
- Clear Structure: Items with visible construction details
- Specific Angles: "90 degrees to the left" vs "turn sideways"
3. Multi-Image Combinations
Combine 6+ different images into one realistic scene:
Example Prompt:
"Combine these images into one realistic photo: the guy holding the peach also holding a basket of peaches at a picnic table. Make it a midday photo at the beach instead of sunset for better lighting."What Works Well:
- Product preservation (logos, colors, shapes)
- Automatic shadow generation
- Lighting consistency
- Realistic placement
Limitations:
- Sometimes misses specific people
- May alter some products slightly
- Complex scenes can lose some elements
4. Person Modification
Edit people while preserving their identity:
Simple Edits:
- "Make the photo warmer and add a mustache"
- "Change the background to a 90s style diner"
- "Add sunglasses and a hat"
Advanced Considerations:
- Identity Preservation: Works well for simple modifications
- Complex Changes: May start to look like a "long-lost brother"
- Environmental Changes: Better to work from original image each time
5. Location Recreation from Maps
Generate realistic views from map locations:
Process:
- Screenshot Map: Google Maps location
- Mark Position: Draw arrow showing viewpoint
- Request View: "What would someone see standing at the tip of the red arrow?"
Results Vary:
- Famous Landmarks: Often accurate
- Street Views: Better with more data available
- Remote Locations: May hallucinate completely new scenes
When to Use Nano Banana vs Photoshop
Use Nano Banana For:
- Quick product mockups
- Natural object placement
- Lighting/shadow matching
- Creative ideation
- Multiple image combinations
- Photo restoration
Use Photoshop For:
- Precise control needed
- Specific positioning
- Text manipulation
- Fine detail work
- Professional retouching
- Brand consistency
Hybrid Workflow
The most effective approach combines both tools:
- Start with Nano Banana: Quick concept and base composition
- Download Result: Get 80% of the way there instantly
- Finish in Photoshop: Fine-tune details and precision elements
- Final Polish: Professional retouching and brand compliance
🚀 The Future of Image Editing
Nano Banana represents a huge leap forward in AI image editing. While it may not completely replace Photoshop for professionals, it democratizes advanced image manipulation for everyone.
Table of Contents
Developer Tutorial
Quick Start Guide: How to Use Nano Banana
Ready to get started with Nano Banana? Follow these simple steps to begin creating amazing AI images and videos:
Choose Your Learning Path
Select the tutorial that matches your needs - beginner guide, developer integration, video creation, or advanced Photoshop alternative techniques.
Access Nano Banana
Go to Google Gemini (gemini.google.com) or AI Studio (ai.studio) and select the Nano Banana (Gemini 2.5 Flash Image) model to start creating.
Start Creating
Follow the step-by-step instructions from our tutorials and start generating amazing images, editing photos, and creating videos with AI.