Taking the parallel machines tools with two redundant drive as the study object, the kinematics model was established, the Jacobian matrix of the machine tool was obtained, and the dexterity of the machine tool was analyzed. Jacobian matrix was used to establish three dexterity indicators of the machine, namely manipulation,the number of conditions and the minimum singular value.Through the three dexterity indicators, the evaluation index comprehensive dexterity coefficient was defined.MATLAB was used to make simulation analysis. The distribution of dexterity at different positions is obtained, which lays a theoretical foundation for machine tool optimization and trajectory planning.