Music plays a vital role in our lives. There is a huge amount of information available everywhere; our task is to filter out only the information that is relevant for users in which they are interested or which suits their goals. In this work, we are aiming to build a music recommendation system that will be content-based. A recommendation system is a filtering system that will predict the user’s choices in music based on their preferences. The content filtering method relies on the characteristics of data and filters it according to the same. So, we implemented a model that works well for a large number of instances and datasets. The design and implementation of the proposed song recommendation system are highly effective and consistently stable. The encouraging results, with an accuracy of 90.5%, provide a new idea and a theoretical basis and open a new avenue for future research on song recommendation systems.