This research paper presents a new approach in which personalized music recommendations are made by combining Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks to form a hybrid model. The study is intended to prove that the model can give accurate song suggestions according to the user’s needs, with data sourced from Spotify through Kaggle platform. Our hybrid CNN-LSTM model has an accuracy of 98.7% which beat traditional recommendation algorithms like CNN, LSTM, Multilayer Perceptron (MLP), Radial Basis Functional Network (RBFN) and Generative Adversarial Network (GAN). It basically shows how an advanced recommendation system can be introduced into a simple application developed using the Streamlit Python framework. Users interact with this application by giving it certain queries to which they will receive personal music proposals. It not only confirms that the hybrid way is effective for improving user engagement and satisfaction on digital music platforms but also signifies an important step towards music recommender systems as it was previously known.