Durian has become a important high-value crop in the Philippines, not only due to local demand but also because it has a high market potential. The Bureau of Agriculture and Fisheries Product Standard(BAFPS) introduced a standardized grading system for fresh durian fruit. However, the current practice of manual grading post-harvest often results in inconsistent classification and inaccurate judgments due to variations among humans. This study aims to design and develop an automated durian-grading machine based on its weight and external characteristics to enhance efficiency and consistency in the grading process compared to the traditional manual approach. The researchers utilized Roboflow as an image processing tool where datasets undergo several processes. The researcher used a confusion matrix to analyze the data gathered. The overall results showed 83.5% accuracy in determining its grade.