The advancement of technology in the field of image recognition has significantly facilitated and improved the effectiveness of object detection in computer-based banknote recognition systems. This study aims to automatically identify banknotes based on their denominations, with the objective of minimizing human errors—such as lack of concentration, fatigue, and other factors—and enabling its application in ATMs and automated payment systems. This research compares the accuracy levels and detection success rates between the YOLO and SSD algorithms in recognizing the denominations of banknotes. The YOLO model operates by dividing the image into grids and predicting bounding boxes along with object classes in a single step, resulting in fast and consistent detection. In contrast, the SSD model employs a multi-scale approach by utilizing feature maps from multiple levels to generate predictions. The parameters used in this study include 7 classes of Indonesian banknotes: Rp1,000, Rp2,000, Rp5,000, Rp10,000, Rp20,000, Rp50,000, and Rp100,000. A total of 353 images were used in the dataset, and three images from each class were selected for testing purposes. The results of the study indicate a significant performance difference. The YOLO algorithm achieved a 100% accuracy rate under both normal and low-light conditions, while the SSD algorithm achieved an accuracy rate of 87.2% under normal lighting and 91.4% under low-light conditions.