Eye diseases, including but not limited to Cataract, Glaucoma and Diabetic Retinopathy, pose significant threats to vision and eye health. This research study underscores the importance of recognizing the profound impact of these conditions on patients’ well-being and seeks to elevate detection methodologies through the application of cutting-edge technologies, particularly deep learning models. The primary objective of this study is to cultivate a model that outperforms existing classification methods, contributing to a more nuanced understanding of eye diseases and facilitating precise and timely detection. In the course of rigorous experimentation and optimization, ResNet-18, a convolutional neural network, emerged as a particularly efficient model. It demonstrated an impressive training accuracy of 96.5% and testing accuracy of 93.7% on a publicly available dataset comprising retinal images of various eye diseases. Such advancements are crucial in addressing the challenges posed by the aforementioned eye diseases, ultimately improving patient outcomes and fortifying initiatives to combat vision-related health issues.