Edge AI combines edge computing and artificial intelligence to help machines make accurate decisions using data from local sources. The effectiveness of AI at the edge is due to advanced computing frameworks like Deep Neural Networks (DNNs) and improved computing infrastructure. However, deploying DNNs on edge devices can be challenging due to their high memory and computational requirements. DNN compression techniques reduce the memory and bandwidth needed for DNNs without sacrificing accuracy, making them suitable for edge devices. Different compression techniques and devices are discussed in the chapter to accelerate AI models at the edge.