Snap a photo of any dish and our Vision AI instantly returns the full recipe — ingredients with quantities, step-by-step instructions, cooking times, and chef tips. Compare two dishes side by side.
Drop a food photo here
Works with any dish — home-cooked, restaurant, street food
Dish → Recipe Decoder uses advanced Vision AI to analyze food photos and return complete, accurate recipes. Whether you're at a restaurant wanting to recreate a dish at home, scrolling through social media and craving something you've seen, or looking at a dish at a dinner party and wanting to learn how it's made — snap a photo and get the full recipe in seconds.
The AI has been trained on millions of dishes from cuisines around the world. It identifies not just the dish name but also the specific preparation style, regional variation, and cooking technique evident in the photo. A plate of pasta triggers identification of the specific pasta shape, likely sauce composition based on color and texture, and the expected ingredients. A grilled meat dish gets analyzed for cut, doneness, and marinade characteristics.
The returned recipe includes ingredient list with quantities, step-by-step instructions, estimated prep and cook times, difficulty level, serving size, chef tips, and a nutritional overview. Results are best for dishes with clear, close-up photography showing the food prominently.