Hybrid welding is a process apply in automotive materials to obtain benefits such as complete penetration, narrow heat affected zones and reduce filler material used. Recently a new industrial revolution is taken place called manufacturing 4.0, the automotive industry has a great interest in all the pillars which making it up, one of the principal pillars is the big data analysis such as welding parameters applied in advances welding processes. The present work describes a welding optimization applying metaheuristic algorithms to find and predict depth penetration (DP) in ASTM 1520 steel welded by a hybrid laser arc welding with the objective to improve the weld quality. Diverse experiments were used to obtain a suitable model, considering welding speed (WS), voltage (V), current (A) and laser power (P). The experimental results demonstrated that the Pareto front values obtained by algorithms improve the DP.