This article presents a study of the performance of the crow algorithm, applied to the control of the Maximum Power Point (MPPT) of photovoltaic panels. The performance of this algorithm was optimized by relying on a meticulous search approach in the promising area. This strategy consists of refining the selection of parameters and the number of search agents and their initial positions to maximize the efficiency of the controller. The proposed algorithm was evaluated by comparing it with the PSO algorithm in terms of efficiency, convergence time and reliability under homogeneous illumination and partial shading conditions. To evaluate the reliability of the controllers, the tests were repeated 60 times, and the standard deviation was used as a measurement indicator. The simulation results in all scenarios (homogeneous illumination and partial shading) validate the superior performance of CSA compared to the PSO technique. This superiority is evident in terms of efficiency, convergence time and reliability.