The Firefly Algorithm (FA) is a highly efficient population-based optimization technique developed by mimicking the flashing behavior of fireflies when mating. This article proposes a method based on Differential Evolution (DE)/current-to-best/1 for enhancing the FA's movement process. The proposed modification increases the global search ability and the convergence rates while maintaining a balance between exploration and exploitation by deploying the global best solution. However, employing the best solution can lead to premature algorithm convergence, but this study handles this issue using a loop adjacent to the algorithm's main loop. Additionally, the suggested algorithm’s sensitivity to the alpha parameter is reduced compared to the original FA. The GbFA surpasses both the original and five-version of enhanced FAs in finding the optimal solution to 30 CEC2014 real parameter benchmark problems with all selected alpha values. Additionally, the CEC 2017 benchmark functions and the eight engineering optimization challenges are also utilized to evaluate GbFA’s efficacy and robustness on real-world problems against several enhanced algorithms. In all cases, GbFA provides the optimal result compared to other methods.