Introduction. In the actual era, the integration of decentralized generation in radial distribution networks is becoming important for the reasons of their environmental and economic benefits. Purpose. This paper investigate the optimal size, location and kind of decentralized generation connected in radial distribution networks using a new optimization algorithm namely bald eagle search. Methods. The authors check the optimal allocation of two kinds of decentralized generation the first is operated at unity power factor and the second is operated at 0.95 power factor, a multi-objective functions are minimized based on reduction of voltage deviation index, active and reactive power losses, while taking into consideration several constraints. Results. Simulation results obtained on Standard IEEE-33 bus and IEEE-69 bus radial distribution networks demonstrate the performance and the efficiency of bald eagle search compared with the algorithms existing in literature and radial distribution networks performances are improved in terms of voltage profile and notably active and reactive power losses reduction, decentralized generation operated at 0.95 power factor are more perfect than those operated at unit power factor.