<p>Due to hybridization of the material removal mechanisms of abrasive jet and water jet machining processes, abrasive water jet machining (AWJM) appears as an efficient non-traditional process providing higher productivity, superior surface quality and excellent dimensional accuracy of almost all types of work materials irrespective of their mechanical and thermal properties. Optimization of this hybrid machining process is a challenging task because of involvement of multiple input parameters, contradictory responses and possible interactions between them. In this paper, based on past experimental datasets, two AWJM processes are optimized using five foraging behavior-based metaheuristic algorithms, i.e. dragonfly optimizer, African vultures optimizer, grasshopper optimizer, fruit fly optimizer and bird swarm optimizer, and their optimization performance is compared with respect to solution accuracy and variability, and computational effort. For both the processes, African vultures optimizer emerges out as the most efficient metaheuristic. For the first example (AWJM of Lanthanum phosphate/Yttria composites), it provides 71.47, 43.73 and 10% improvements for single-objective optimization; and 68.37, 43.45 and 9.53% improvements for multi-objective optimization, in material removal rate, kerf angle and surface roughness, respectively against the observations of the past researchers. On the other hand, in case of the second example (AWJM of glass fibre-reinforced polymer composites), the corresponding improvements are 41.62, 22.35 and 27.90% for single-objective optimization; and 23.66, 9.09 and 28.57% for multi-objective optimization in surface roughness, kerf taper and delamination length, respectively. Based on the Friedman’s mean rank test, it can also be noticed that African vultures optimizer supersedes the remaining foraging behavior-based metaheuristic algorithms.</p>