Generative AI, often known as genAI, encompasses several forms of artificial intelligence (AI) that has the ability to create unique text, images, video, or audio content. This particular iteration of artificial intelligence acquires knowledge of patterns and data arrangement from its training data, enabling it to produce novel outputs that possess similar statistical characteristics. Generative AI has a diverse range of applications, and each task requires a specialized deep-learning architecture to effectively capture the unique patterns and traits found in the training data. Generative AI models encompass various types, including generative adversarial networks (GANs), variational autoencoders (VAEs), transformers, diffusion models, normalizing flow models, and hybrid models. The configuration of a generative AI model is contingent upon the particular task and domain, encompassing elements such as the neural network's architecture, training approach, loss function, and evaluation metrics. The primary objective of generative AI is to develop autonomous systems capable of generating content that is indiscernible from information created by humans. This encompasses the production of written content, visual graphics, audio recordings, and interactive visual components. Attaining this objective would facilitate a diverse array of applications, encompassing enhanced human-computer interactions and assisting in the advancement of endeavors such as art and storytelling.