This paper focuses on the finite-time output containment problem for a kind of nonlinear multi-agent systems with multiple dynamic leaders. Firstly, considering the topological structure among the followers, a kind of adaptive distributed observer is designed to estimate the whole states of all the leaders. By utilizing common Lyapunov theory, the finite-time convergence of proposed distributed observer is proved. On the basis of this conclusion, a containment control protocol including the desired convex combinations of the leaders is developed for each follower by using the given weights. With the help of the output regulation theory, the finite-time output containment criterion for nonlinear multi-agent systems is derived. Finally, a numerical example is presented to demonstrate the effectiveness of the theoretical results.