Diabetic retinopathy (DR) stands as a formidable threat to global vision health, representing a leading cause of blindness. Early detection of this ailment plays a pivotal role in mitigating its impact by enabling timely intervention and treatment, thereby reducing the risk of vision loss. In response to this critical healthcare challenge, our study endeavors to develop a Deep Learning-based model designed to assess an individual’s susceptibility to Diabetic Retinopathy. The proposed model employs hybrid model of Convolutional Neural Network (CNN) and Inception ResNet V2 to analyze eye fundus images which classify them into one of four distinctive categories: non-DR, mild DR, moderate DR, and severe DR, with an additional classification for proliferative DR (PDR). Extracting a dataset comprising of 3662 images from 5590 images to ascertain its efficacy and performance, contributing significantly to the realm of preventative healthcare and vision preservation.