🧠 Snackhack 🍳

Snackhack is a fullstack AI-powered fridge recipe finder app. Just enter ingredients you have in your fridge, and it generates delicious recipes with realistic images using GPT-4 and DALL·E.

Built with:

• ⚛️ React + Vite + Tailwind CSS (frontend)
• 🐍 Flask (backend)
• 🤖 OpenAI API (GPT-4 + DALL·E for images)

---

🚀 Features

• Ingredient input with tag-style autocomplete
• Filtering (Vegetarian, Under 30 Min, etc.)
• Dark/light mode toggle
• Like & save recipes (locally stored)
• GPT-powered recipe generation
• AI-generated food images (via DALL·E 2)

---

🛠 Getting Started

🔧 Prerequisites

• Node.js (v18+ recommended)
• Python 3.10+
• An OpenAI API Key

---

📦 Setup

1. Clone the repo

git clone https://github.com/your-username/snackhack.git
cd snackhack

---

2. Frontend (React + Vite)

cd frontend
npm install
npm run dev

Frontend will run at: http://localhost:5173

---

3. Backend (Flask + OpenAI)

cd backend
python -m venv venv
source venv/bin/activate  # or venv\Scripts\activate on Windows
pip install -r requirements.txt

Create a .env file in the backend/ folder:

OPENAI_API_KEY=your-openai-key-here

Then run the server:

python app.py

Backend runs on http://localhost:5000

---

🧪 Example Usage

1. Go to the frontend in your browser.
2. Add ingredients like: chicken, pasta, cheese
3. Click Cook.
4. Wait ~5–10s (image generation takes time).
5. Scroll through generated recipes.

---

💸 API Notes

• DALL·E 2 is used for images (cheaper + 512x512 support)
• API rate limit: 5 images/min (use fallback images if needed)
• GPT-4 or 3.5 Turbo used for recipe generation

---

📁 File Structure

/frontend         → React + Tailwind UI
/backend          → Flask API & GPT logic
/images           → Optional logos & assets
README.md
.gitignore

---

✨ Future Improvements

• User login & recipe saving
• Mobile PWA support
• Multiple image quality options
• Upload your own fridge photo

---

🧑‍💻 Made by Nikola Ilić

Junior Front-End Developer | React · Flask · Tailwind