Mobile video traffic accounts for more than half of the global mobile data traffic nowadays, and the ratio is expected to further increase in the near future. However, providing high quality of experience for video streaming in mobile networks is challenging due to the heterogeneous and varying wireless channel conditions. To meet the increasing demand of high-quality mobile video streaming services, researchers have proposed several cooperative video streaming models that enable mobile users to download video contents cooperatively. The key idea is to pool network edge resources so as to either alleviate the load on the video servers and the cellular network, or alleviate the impact of channel variations and improve resource utilization. In this article, we review four types of cooperative video streaming models that pool various network resources effectively in different application scenarios. Then we focus on the crowdsourced mobile streaming model, which aims to pool users' download capacities in order to alleviate the impact of channel variations and achieve efficient utilization of network resources. We introduce the corresponding optimization issue of efficient resource allocation and the economic issue of user cooperation. We also outline future challenges and open issues in cooperative video streaming models.