Our research delves into enhancing user experiences by introducing mood-based music recommendations without the need for intrusive wearable devices. Through natural language processing (NLP), we successfully classified user-entered text into five emotional categories with a commendable 74.22% accuracy which is later used to recommend appropriate songs to the user. This methodology not only revolutionizes personalized music curation but also extends its applications to broader realms, such as assessing overall emotional wellbeing. Our non-intrusive and easily implementable approach offers a scalable solution, providing a seamless and meaningful way for applications to understand and respond to users’ emotions in real-time, transcending the limitations of current solutions.