Deep learning has revolutionized various fields by enabling machines to learn complex patterns from data. However, a significant limitation of traditional deep learning models lies in their focus on classification or prediction tasks. Generative models, a subfield of deep learning, address this gap by learning the underlying distribution of data and generating entirely new samples that resemble the training data. This article delves into the concept of generative models, exploring their core principles and the various techniques employed, with a particular focus on Generative Adversarial Networks (GANs). We discuss the applications of generative models in creating realistic images, sounds, and even text formats, highlighting their potential to revolutionize various creative industries and scientific research fields.