This study proposes a hybrid SqueezeNet-vision transformer (SViT) model for automated retinal disease classification using optical coherence tomography (OCT) images. The model combines the strengths of SqueezeNet and vision transformers to capture both local and global features of OCT images, aiming to achieve more accurate classification with reduced computational complexity. The SViT model is trained on the OCT2017 dataset and is capable of performing both binary (normal vs. disorders) and multiclass (DME, CNV, Drusen, and normal) classifications. Compared to state-of-the-art CNN-based and standalone Transformer models, the SViT model achieves an overall classification accuracy of 99.90% for multiclass classification. The model's good generalization ability makes it suitable for improving patient care and clinical decision-making in ophthalmology. The study highlights the importance of early detection and diagnosis of retinal diseases and the role of AI methods in enhancing medical image analysis.This study proposes a hybrid SqueezeNet-vision transformer (SViT) model for automated retinal disease classification using optical coherence tomography (OCT) images. The model combines the strengths of SqueezeNet and vision transformers to capture both local and global features of OCT images, aiming to achieve more accurate classification with reduced computational complexity. The SViT model is trained on the OCT2017 dataset and is capable of performing both binary (normal vs. disorders) and multiclass (DME, CNV, Drusen, and normal) classifications. Compared to state-of-the-art CNN-based and standalone Transformer models, the SViT model achieves an overall classification accuracy of 99.90% for multiclass classification. The model's good generalization ability makes it suitable for improving patient care and clinical decision-making in ophthalmology. The study highlights the importance of early detection and diagnosis of retinal diseases and the role of AI methods in enhancing medical image analysis.