This document presents a neural network model for classifying audio files into genres based on their acoustic features. The model was trained on a dataset of 1000 audio clips from 10 genres. It uses a 5-layer neural network that takes features like MFCCs and chroma coefficients as input and outputs the predicted genre. The model achieved over 70% accuracy on a test set of 200 audio clips, correctly classifying genres for about 7 out of every 10 clips. While the model provides a faster way to classify audio without human input, limitations include only classifying genres it was trained on and potential for misclassification of complex audio tracks.