Video models extend AI beyond static images by enabling systems to understand and generate dynamic visual content over time. They can analyze sequences of frames to detect actions, track objects, summarize videos, and answer questions about events happening in a scene.

At the same time, generative models can create entirely new videos from text prompts, images, or scripts. These capabilities are powering applications such as video search, automated editing, content creation, surveillance analysis, and multimodal assistants.

In this chapter, you will learn how video understanding and generation models work and how to use them to build real-world AI applications.