The performance and efficiency of twelve range-separated hybrid functionals are compared with each other in calculating magnetic exchange coupling constant (J) data on eleven di-nuclear first-row transition metal complexes having di-copper (Cu) and di-vanadium (V) as their central atoms. The J values are then compared with the experimental findings. In addition to that we also computed the B3LYP results of these complexes. We compared them with the results obtained from range-separated functionals as well as with the experimental results to understand their performances in predicting the magnetic exchange coupling constant. When computing, we reoptimized all the structures of the complexes taken from the crystal data, and then the values of Js were determined. The performances are predicted in terms of four statistical error matrices, namely, the mean absolute error (MAE), mean fractional error (MFE), mean signed error (MSE), root mean square error (RMSE) in cm−1.