The Question Paper Prediction Model for Examination is designed to assist students in preparing for exams more effectively by recommending the most relevant questions based on historical data. The system processes previous years’ question papers using HDBSCAN (Hierarchical Density-Based Spatial Clustering of Applications with Noise), an advanced clustering algorithm. It leverages Natural Language Processing (NLP) techniques to clean the question data, generates embeddings using pre-trained Sentence Transformers, and clusters similar questions to identify important patterns. The model continuously integrates new question papers to provide updated, relevant recommendations. By reducing exam stress and uncertainty, the model supports better mental well-being and more focused study.