This paper investigates the finite-time group consensus problems for networked multi-Lagrange systems （MLSs）with parameter uncertainty under directed topology. To address the problems of nonlinear Lagrange dynamics with parameter uncertainty，the paper proposes adaptive finite-time control protocols by introducing two types of sliding vectors. The results show that the given sliding control in the paper solves the nonlinear problems of the MLSs under the group networks and ensures the finite-time consistent convergence of the system. Under the group topology，the networked MLSs are distributed to achieve group finite-time consensus. Finally，the simulation results illustrate the feasibility of the control protocol，and show the effectiveness of the finite-time control methods.