Currently, technological development is progressing rapidly. One of the areas is in artificial intelligence and computer vision technology. One of the branches of science studied in artificial intelligence and computer vision technology is deep learning, which focuses on the use of artificial neural networks that learn about classification and object detection directly through images and videos. With the advent of deep learning, researchers are focusing on developing a jengkol (dogfruit) ripeness detection system using deep learning methods. This research uses 1500 jengkol images as a dataset and uses Roboflow for labeling. The results of the labeling will be divided into 3 types of classes: young, medium, and old jengkol. The YOLOV5 algorithm is used for training the jengkol dataset. The next stage is testing, where the approaches used are confusion matrix, classification report, Mean Squared Error (MSE), and Root Mean Squared Error (RMSE), as well as the F1-Score value. The purpose of this testing is to see the precision results and identify the optimal accuracy on data from the trained model that can be achieved by the system being tested or evaluated, which involves compiling a classification report into three categories: young, medium, and old, based on visual characteristics that can be detected by the deep learning algorithm. From the real-time testing and evaluation results obtained in this study, the accuracy value is 80%.