Disease effects in case of apples are not negligible and pose a real challenge to agricultural output and nations’ well-being. This study applies deep learning methods in classifying the illnesses that impact the leaves of the apple tree with the aim of boosting the techniques of organic farming. As in prior studies, EfficientNetB3 model, which is both efficient and less time-consuming, was used to build a dedicated model for apple leaf’s diseases detection. The mock-apple tree data set which is known as Apple Leaves Disease Dataset (ALDD) was used in training and testing the model and the photos present in the dataset is more than 9700 and classified under the mock-apple healthy leaves and other diseases. That kind of technique we have used was very accurate and specific since the model we developed was able to show the existence of Apple Scab, Black Rot, Cedar Apple Rust and even the healthy leaves with a 100% detection rate. Precision obtained the score of 1 for the identification of health leaf, Apple Scab, Black Rot, and Recall also attained the score of 1 for the identification of health leaf, Apple Scab and Black Rot, and F1-score also got the highest score of 1 for the identification of the health leaf, Apple Scab, Black Tot. Nevertheless, for the current pattern Cedar Apple Rust got a little lower score, 99%, for both the criteria that refer to Recall and F1-Score. The results of employing the described approach to diagnosis of diseases in apple trees prove rather high effectiveness of the approach in identification of illnesses in early stages, which is a crucial factor in managing diseases affecting apple leaves, and the scalability of the approach. This study can be important to the sustainable agriculture problems in apple growing by applying superior deep learning algorithms. This is done as a result of minimizing the degree of the chemical treatment and hence enhances control of diseases affecting plants.