In this project, we develop a music mood classification model to classify the mood of the music into 6 categories: Anxious, Chill, Focus, Party, Sad, and Romance. The music is preprocessed in order to use as the music database for generating exercise music tracks based on the user's requirements. To keep it easy and simple, we also develop "Melodistic" which is a mobile application for streaming exercise music tracks and generating new tracks based on mood. It also allows the user to include their own favorite music.