The detection and classification of counterfeit currency is a critical task for banks, businesses, and individuals alike. In this paper, we present a deep learning approach for identifying fake and genuine currency using a dataset of 92,000 images. Our dataset includes 16 classes, eight of which are faked and eight of which are genuine. We employ a convolutional neural network architecture to train our model on this data and achieve a high level of accuracy in detecting counterfeit currency. Specifically, our model achieves an accuracy of 98% on the testing dataset. We also compare our results to previous approaches and demonstrate that our deep learning approach outperforms these methods. Overall, our work contributes to the development of automated systems for detecting counterfeit currency and has important implications for security and financial fraud prevention.