The world of AI is has gone through tremendous advancements in over last decade. AI journey started from Algorithmic Intelligence and Linguistic Intelligence, and is moving towards Imaginative intelligence. This generative AI’s historical journey is focused on generating realistic texts and images through the means of early models like autoencoders and variational autoencoders (VAEs). From creating simple and coherent images from basic textual descriptions using Generative Adversarial Networks (GANs) and models like DALL-E in text to image generation, the AI leaped towards generating coherent frames which aligned for improving video quality, length, and the ability to generate more complex scenes in video with natural transitions in the field of text to video generation. The T2V models made use of models such as diffusion model, transformer model and the latest and the best one which is diffusion transformer model (DiT model). This paper basically gives us the overview of text to image and text to video generation models and their advancement in the field of generative AI.