The paper provides a comprehensive review of the firefly algorithm (FA), a nature-inspired meta-heuristic optimization method. FA is inspired by the flashing lights of fireflies and their behavior, which includes attraction and repulsion based on light intensity and distance. The algorithm is designed to solve various optimization problems, including continuous, combinatorial, constrained, and multi-objective optimization, as well as dynamic and noisy environments. The paper discusses the biological foundations of FA, its algorithmic structure, and its key characteristics, such as the variation of light intensity and attractiveness. It also reviews the modifications and hybridizations of FA, highlighting improvements in areas like elitism, binary representation, Gaussian distribution, Lévy flights, and parallelization. The paper further explores the applications of FA in engineering, machine learning, data mining, and neural networks, demonstrating its effectiveness in solving real-world problems. The review concludes by discussing the efficiency of FA and its potential for further development in swarm intelligence.The paper provides a comprehensive review of the firefly algorithm (FA), a nature-inspired meta-heuristic optimization method. FA is inspired by the flashing lights of fireflies and their behavior, which includes attraction and repulsion based on light intensity and distance. The algorithm is designed to solve various optimization problems, including continuous, combinatorial, constrained, and multi-objective optimization, as well as dynamic and noisy environments. The paper discusses the biological foundations of FA, its algorithmic structure, and its key characteristics, such as the variation of light intensity and attractiveness. It also reviews the modifications and hybridizations of FA, highlighting improvements in areas like elitism, binary representation, Gaussian distribution, Lévy flights, and parallelization. The paper further explores the applications of FA in engineering, machine learning, data mining, and neural networks, demonstrating its effectiveness in solving real-world problems. The review concludes by discussing the efficiency of FA and its potential for further development in swarm intelligence.