Aiming at the tension distribution redundancy problem in tendon-driven dexterous hands, this study proposes an optimal tension distribution method based on quadratic programming to achieve hybrid force-position control. By establishing the kinematic and static models of the dexterous hand’s fingers and designing a tension optimization algorithm, the desired tendon forces are derived by integrating position and force control objectives, ultimately realizing hybrid force-position control for the fingers. Simulation results demonstrate that this method enables real-time tendon tension distribution, accurately tracking both desired contact forces and joint positions, making it suitable for highly dynamic manipulation tasks in dexterous hands.