This paper proposes a vortex search algorithm (VSA) optimization for optimal dimensioning of distributed generators (DGs), in radial alternating current (AC) distribution networks. The VSA corresponds to a metaheuristic optimization technique that works in the continuous domain, to solve nonlinear, non-convex, large scale optimization problems. Here, this technique is used to determine the optimal power generation capacity of the DGs from the top-down analysis. From the bottom-up, a conventional backward/forward power flow is employed for determining the voltage behavior and calculate the power losses of the network, for each power output combination in the DGs. Numerical results demonstrate that the proposed approach is efficient and robust for reducing power losses on AC grids by optimally sizing the capacity the DGs, compared with other approaches found on literature reports. All the simulations were conducted using the MATLAB software.