Summary In recent years, reactive power compensation techniques are receiving more attention across the globe from both the distribution network operators and research groups due to its ability in the reduction of network power loss, voltage deviation, and enhancement of voltage stability. In this paper, two novel and robust optimization techniques implemented for solving optimal capacitor (OC) allocation problem in various distribution networks are water cycle algorithm (WCA) and grey wolf optimizer (GWO). To crisscross the feasibility, GWO and WCA techniques are tested on practical Indian 28‐bus, 47‐bus, and 52‐bus and standard 33‐bus, 69‐bus, and 85‐bus radial distribution networks with uniformly varying load levels. The simulation results obtained by WCA and GWO are compared with other techniques like inclusion and interchange variable (IIV), direct search algorithm (DSA), and teaching and learning–based optimization (TLBO) and found to be better. In addition, from the obtained results, the algorithm's efficiency is evaluated in terms of exploration and exploitation, and convergence characteristics and computational time are discussed.