Stock Image Inspirations

A blackbox FAL serverless service that applies professional photo editing inspirations to images using Google's Nano Banana model.

Features

• 🎨 8 Professional Inspirations - Variations, marketplace, cinematic, and more
• 🖼️ Always 3 Images - Consistent output for every request
• 📐 Aspect Ratio Support - 1:1, 16:9, 4:3, and more
• ⚡ Fast & Scalable - CPU-optimized, scales to zero
• 🔒 Blackbox Design - Simple API, complex prompts handled internally

Live Endpoint

Playground: https://fal.ai/models/Adc/stock-inspirations/

API:

• Async (Recommended): https://queue.fal.run/Adc/stock-inspirations/
• Sync: https://fal.run/Adc/stock-inspirations/

Quick Start

1. Install Dependencies

pip install fal-client python-dotenv

2. Setup Environment

Create .env file:

FAL_KEY=your-fal-api-key-here

3. Use the API

import fal_client

# Upload your image
image_url = fal_client.upload_file("your_image.jpg")

# Apply inspiration
handler = await fal_client.submit_async(
    "Adc/stock-inspirations",
    arguments={
        "inspiration_name": "marketplace_pure",
        "image_urls": [image_url],
        "aspect_ratio": "1:1",  # Optional
        "extra_prompt": "vibrant colors"  # Optional
    }
)

result = await handler.get()

# Get your 3 generated images
for img in result['images']:
    print(img['url'])

Available Inspirations

Inspiration	Description	Images Required
variations	Professional variations with different styles	1
marketplace_pure	Clean product photography (white background)	1
marketplace_lifestyle	Lifestyle marketplace photography with context	1
change_pose	Change subject pose while maintaining identity	1
style_cinematic	Apply cinematic film photography style	1
background_white	Replace background with white studio background	1
enhance	Enhance image quality and sharpness	1
fuse_images	Combine multiple images into compositions	2-5

Aspect Ratios

Supported aspect ratios (optional):

• 21:9 - Ultra-wide
• 16:9 - Widescreen
• 4:3 - Standard
• 3:2 - Classic photo
• 1:1 - Square
• 9:16 - Vertical
• And more...

Input Parameters

{
    "inspiration_name": str,      # Required: Name of inspiration
    "image_urls": List[str],      # Required: List of image URLs
    "aspect_ratio": str,          # Optional: Output aspect ratio
    "extra_prompt": str           # Optional: Additional instructions
}

Output Format

{
    "success": bool,
    "images": [
        {"url": str, "index": int},  # 3 images
        {"url": str, "index": int},
        {"url": str, "index": int}
    ],
    "inspiration_name": str,
    "prompt_used": str,
    "aspect_ratio": str,
    "processing_time": float,
    "request_id": str,
    "error": str                    # Only if success=False
}

Examples

Run examples:

python example_usage.py

Test deployed endpoint:

python test_deployed_endpoint.py

Development

Deploy to FAL

# Authenticate
fal auth login

# Deploy
bash deploy.sh
# or
fal deploy stock_inspirations_app.py

Local Development

# Create virtual environment
python -m venv venv
source venv/bin/activate  # or: venv\Scripts\activate on Windows

# Install dependencies
pip install -r requirements.txt

# Run examples
python example_usage.py

Architecture

• Machine Type: M (CPU) - Fast deployment, low cost
• Concurrency: 0-2 workers, scales to zero
• Timeout: 120s per request
• Engine: Google Nano Banana Edit model via FAL

Cost

Typical processing time: 9-11 seconds per request
Cost: Based on FAL's nano-banana/edit pricing

License

MIT