Date palm production is vital to agriculture in arid regions, but diseases can severely affect yield and quality. Early detection is essential for effective disease management. This study proposes a Vision Transformer-based framework for automatic date palm disease identification from crop images, enhanced with explainable AI techniques to provide insights into the model’s decision-making process. The model is trained on a dataset of disease symptoms like brown spots, white scales, and healthy palm conditions using RGB images. Performance is evaluated using accuracy, precision, recall, and F1-score metrics. Results show that the Vision Transformer outperforms traditional CNNs’ accuracy and sensitivity to subtle disease signs. Explainable AI methods further improve interpretability, allowing users to understand the reasoning behind predictions. This framework offers a promising tool for real-time disease monitoring in date palm plantations, advancing smart agriculture applications and enabling better-informed decision-making.