Experimental bulk mixing data on disordered bimetallics of Ni, Cu, Ph, Pd, Ag, Ir, Pt, and Au are used to parametrize the recently developed bond order metal simulator (BOS-mixing) model, including a full error analysis. This model characterizes the variation of metal-metal bond strength with number and type of atomic neighbors. The model is shown to accurately ht experimental mixing energy curves as a function of composition irrespective of whether the curves are symmetric or asymmetric around the 50% value. As an illustration of the utility of the BOS-mixing model, we predict the microstructures of bimetallic clusters with 201 atoms (or 61% dispersion) and a composition of 50%-50%. The examples demonstrate how differences in surface energy, mixing energy, and mixing entropy either compete or cooperate in determining the microstructure of small bimetallic clusters.