With the continuous development of multimodal large model technology, it has been widely applied across various domains, demonstrating excellent performance in specific tasks such as visual question answering, AIGC, and embodied intelligence. As the capabilities of multimodal large models improve, their scale also increases, requiring more powerful computational resources and more complex designs. This demand raises the complexity of each model and necessitates more paired data, making model integration challenging. This chapter will introduce typical applications of multimodal large models and review the latest trends in multimodal applications. Additionally, it will discuss the general technical paradigms for applying multimodal large models. More specifically, the author will present three representative applications: visual question answering, AIGC, and embodied intelligence. Finally, some current unresolved issues and challenges will be discussed, along with potential future research directions.