The emergence of Green IoT with deep learning helps the agriculture sector detect different diseases more efficiently. In this paper, a Green IoT-based smart system has been developed using deep learning and mobile technology to accurately detect and classify tea leaf diseases. The work used Green IoT principles to make the system more energy-efficient and environment-friendly assisted by the deep learning process such as convolutional neural network (CNN) to analyze images of tea leaves with a developed easy-to-use mobile application interface for capturing and uploading images of tea leaves through the device for analysis. The work achieved 88% validation accuracy by using the proposed CNN model. The fusion of Green IoT with easy-to-use mobile applications and deep learning process is novel and provides tea farmers as well as agricultural experts with a powerful tool for accurately identifying and managing tea diseases promptly, ultimately improving crop yields and reducing waste.