A new deep learning model, the Attention-driven YOLOv11n–DeiT, has been developed for enhanced detection of tomato leaf diseases, according to a study in Scientific Reports. Tomatoes are highlighted as the most widely consumed and economically significant vegetable crop worldwide, second only to potato in global importance, and are a primary dietary source of lycopene, an antioxidant associated with lower risk of cardiovascular illnesses and some cancers. The tomato plant is distinguished by compound, serrated leaves and glandular trichomes, with high genetic diversity encompassing thirteen closely related species. The industry faces challenges from frequent diseases, which represent a significant barrier to high and stable productivity, and recurrent plant diseases necessitate extensive pesticide use, raising environmental concerns and increasing production costs. Standard diagnostic techniques, such as expert visual assessments and laboratory-based pathogen detection, are time-consuming, labor-intensive, and dependent on personal skills, limiting scalability. Advancements in deep learning, including convolutional neural networks, provide alternatives for plant disease detection by facilitating accurate analysis of complex image data, offering automation, adaptability, and efficient handling of large-scale datasets. The new model aims to enable real-time diagnosis with minimal human intervention, supporting sustainable agricultural practices and crop monitoring.