Fridge to Dish: Your AI Sous-Chef, Running Entirely in Your Browser

Welcome to Fridge to Dish, a cutting-edge, standalone web application that transforms your leftovers into culinary masterpieces. Built using the principles of „Vibe Coding,“ this app leverages the power of modern browser technology and advanced AI to create a fully private, highly scalable, and incredibly fast cooking assistant.

---

🍽️ What is „Fridge to Dish“?

Fridge to Dish is your ultimate kitchen companion. Instead of endlessly scrolling through recipe blogs trying to find something that matches the random ingredients in your refrigerator, you simply tell the app what you have.

Using state-of-the-art Generative AI, the app instantly crafts tailored recipes complete with step-by-step instructions, nutritional calculations, beautiful food imagery, and organized shopping lists for missing items.

Key Features at a Glance:

• Multi-Modal Input: Upload a photo of your fridge, type out your ingredients, or simply speak to the app using built-in Voice Input (utilizing the native Web Speech API).
• Dynamic AI Recipes: Powered by Google Gemini (dynamically discovering and using models like gemini-1.5-flash or gemini-1.5-pro based on your key).
• Stunning Food Photography: Generates realistic images of your customized dishes using Google Imagen 3 (or falls back to Pollinations.ai or an interactive canvas label if needed).
• Smart Shopping Lists & Meal Planning: Generate multiple recipes at once, combine them, and automatically merge quantities and units into a single, organized shopping list.
• Privacy-First History: Save your favorite creations. All search history, including compressed images (to prevent browser storage limits), is saved directly on your device.
• Multi-Language Support: Full translation and generation capabilities in English, German, French, Spanish, and Italian.
• Interactive Video Guide: Features an integrated video avatar of Claude Cristhi, greeting you dynamically depending on your language.

---

🚀 The Standalone Architecture Advantage

Unlike traditional web applications that rely on heavy, expensive cloud backends, Fridge to Dish runs 100% client-side. This architectural decision brings massive benefits to users:

🔒 1. Ultimate Privacy & Security

Your data belongs to you. Because the app runs entirely in your local browser:

• Your Google Gemini API key is stored securely in your browser’s local storage. It is never sent to any third-party server.
• Your photos, ingredients, and recipe history remain on your machine.
• No accounts, no tracking, and no database breaches to worry about.

⚡ 2. Cost-Effective & Scalable

By eliminating the backend server, the app is incredibly lightweight and serverless:

• No subscription fees: You run the app using your own API key, meaning you only pay Google directly for the tokens you actually use (often free or mere fractions of a cent per request).
• Infinite scaling: Since every user’s browser does the heavy lifting, the website will never slow down or crash due to high server load.

💾 3. Smart Local Data Management

Browsers have storage limits, but Fridge to Dish is engineered to handle them gracefully:

• Dual-Database Fallback: The app automatically prefers IndexedDB for robust local storage. If that fails (e.g., when run locally via file://), it silently falls back to localStorage.
• Automatic Image Compression: To prevent QuotaExceededError in local storage, high-res AI images are scaled down and compressed on an invisible HTML5 canvas before saving.

🌐 4. Zero Installation & Fully Portable

Since the entire application is bundled into a single file (f2d.html), you can run it anywhere:

• Host it easily on a standard WordPress site by uploading it to the root folder.
• Open it locally on your computer or phone without an internet connection (as long as you have API access).
• No complex installations, node servers, or Docker containers required.

---

[!TIP]
Why „Vibe Coding“?
Fridge to Dish is a prime example of modern AI-assisted development. By focusing on clean, standalone HTML5/JS architecture and delegating core AI features to direct API integration, we created a feature-rich, high-performance app with zero backend overhead.