In today’s world, selecting a proper recipe depending upon the ingredients available in the home is time-consuming and inconvenient. Current systems take a lot of time in asking for ingredient names manually, which lowers efficiency and user satisfaction. To overcome this, we suggest implementing an AI powered recipe suggesting system that automatically identifies ingredients from images and give corresponding recipe suggestions. The two-stage system consists of: (1) Ingredient Recognition, a Convolutional Neural Network (CNN) scans the uploaded images to determine ingredients based on color, and shape; and (2) Recipe Recommendation, a content based filtering algorithm compares the identified ingredients against the appropriate recipes in a database. In contrast with other methods, our system makes it possible for individuals to merely upload a single image or multiple images of the ingredients instead of typing the text manually, thus making the process easier. This method not only saves time and increase efficiency but also minimize food wastage by proper utilization of existing ingredients.