Generate scalable e-commerce product images with GPT-4 and NanoBanana Pro

🚀 AI Image Generation Workflow – Scalable E-commerce Product Images

This workflow automates the creation of high-quality, AI-generated product images using NanoBanana Pro.
It analyzes multiple reference images, generates a professional photoshoot-style prompt, creates a new image, and stores the final result with a public URL for reuse.

📄 Documentation: Notion Guide

👤 Who is this for?

This workflow is designed for:

• E-commerce store owners
• Digital marketers and growth teams
• Creative agencies
• Automation builders using n8n
• Anyone who wants to generate scalable, consistent product images from existing photos

No advanced coding skills are required.

❓ What problem does this workflow solve? / Use case

Creating professional product images at scale is expensive, slow, and inconsistent.

This workflow solves:

• Manual photoshoot costs
• Inconsistent visual branding
• Time wasted on prompt writing
• Difficulty generating AI-ready public image URLs
• Repetitive image upload and storage steps

Typical use case:
Transform 3 reference photos (model + product) into a studio-quality fashion image automatically.

⚙️ What this workflow does

1. Collects exactly 3 images via a form upload
2. Validates inputs to ensure all required images are present
3. Splits images into individual processing paths
4. Uploads original images to Google Drive (permanent storage)
5. Generates public, crawlable image URLs
6. Analyzes each image using AI vision (GPT-4O)
7. Aggregates image descriptions into a structured context
8. Generates a professional photoshoot prompt using an AI agent
9. Creates a new image via NanoBanana Pro
10. Polls the API until the image generation is completed
11. Downloads the final image as a binary file
12. Uploads the final image to Google Drive
13. Logs results (images + descriptions) into Google Sheets

🛠️ Setup

Required credentials

• Google Drive (OAuth)
• Google Sheets (OAuth)
• OpenAI API key
• AtlasCloud API key

Required configuration

1. Replace all <__PLACEHOLDER_VALUE__> fields:
   • Google Drive folder IDs
   • Google Sheets document ID and sheet name
   • AtlasCloud API key
2. Ensure Google Drive folders have write permissions
3. Confirm tmpfiles.org is reachable from your environment

Important notes

• The workflow expects exactly 3 images
• The final image is downloaded as binary before upload
• Public URLs are normalized to https://tmpfiles.org/dl/... for maximum AI compatibility

🎥 Watch This Tutorial

👋 Need help or want to customize this?

📩 Contact: LinkedIn
📺 YouTube: @DRFIRASS
🚀 Workshops: Mes Ateliers n8n

Need help customizing?

Contact me for consulting and support : Linkedin / Youtube / 🚀 Mes Ateliers n8n