It is known that Proportional-Integral-Derivative (PID) controllers and their variations are widely used to control industrial processes, both in single-variable or in multivariable systems, which leads to its use in about 95% of industrial plants. In this context, this work presents the study of Proportional-Integral (PI) controller optimization gains with the anti-windup technique, applied to a level system from PD3 SMAR didactic plant, through Bee Swarm Optimization (BSO). For this, initially, the system open-loop response was obtained, under the application of first-order identification methods. Next, the PI controller was tuned using the BSO optimization algorithm, being compared with the Haalman classical tuning method, only for comparative analysis. For the results, of both controllers, a 40% Set-Point (SP) was performed using the Integral of the Squared Error (ISE) and Integral of the Quadratic Error multiplied by time (ITSE) performance indices for the control loop assessments. It was possible to observe that the PI controller tuned by the BSO algorithm obtained better performance. Therefore, it is worth highlighting how beneficial the optimization algorithm was compared to classical control techniques.