Metaheuristic algorithms are widely used for solving complex optimization problems due to their flexibility and efficiency. Among them, the Black Winged Kite Algorithm (BKA) has shown promising results but suffers from instability, premature convergence, and uneven distribution of the initial population. To address these limitations, this paper proposes an Enhanced Black Winged Kite Algorithm (EBKA), which incorporates Latin Hypercube Sampling (LHS) for better initial diversity and Lévy flights to achieve a balanced exploration–exploitation trade-off. Experimental evaluations on CEC2022 benchmark functions demonstrate that EBKA consistently achieves superior stability, faster convergence, and improved solution quality compared to the original BKA and other recent metaheuristics.