This paper presents an optimal control framework tailored for redundant robotic manipulators, aiming to devise precise joint-space trajectories while minimizing control efforts. The core contribution is the formulation of trajectory planning as a multi-objective optimization problem, tackled through a Genetic Algorithm-based Model Predictive Control strategy, followed by Gradient Descent refinement. Moreover, we develop a strategy to apply the resulted high-level design of the joint-space trajectory into a dynamic time-series control strategy that respects the physical constraints of actuators in motion, speed, acceleration and jerk. Experimental results on a simulation model underscore the framework’s efficacy, demonstrating minimal positioning errors without the need for high computational resources for the trajectory design. Moreover, dynamical analysis of the actuators signals for the low-level phase demonstrates the ability of the overall framework to be applied on a real robotic manipulator.