Rapid discovery of a deep eutectic solvent (DES) and its increasing application in various scientific applications demands a well-defined polarity scale for the new generation of eutectic solvents. The tunable nature of a DES facilitates to increase its number by combining various hydrogen bond acceptors and donors in multiple molar ratios. Therefore, the experimental measurement of polarity scale represented by solvatochromic parameters is a cumbersome task. In this context, we have developed a first-principle-based conductor-like screening model-real solvent (COSMO-RS)-derived predictive modeling approach to predict the Kamlet-Taft solvatochromic parameters of 26 hydrophilic and 62 hydrophobic DESs. The COSMO-RS-derived energies, that is, hydrogen bonding energy (E-HB), misfit energy (E-MF), and van der Waals energy (E-VDW), are linearly correlated with the experimental data set of 28 DESs having 4 choline chloride, 17 tetraalkylammonium organic salt, 4 menthol, and 3 betaine subclasses. A lowest root-mean-square deviation (rmsd) of 0.06% for the choline chloride class of DESs and a highest rmsd of 3.14% for the tetrabutylammonium chloride class of DESs are obtained. Therefore, the correlated models are then extended to develop 60 DESs consisting of 15 choline chloride DESs, 4 different hydrophilic DESs, 17 tetraalkylammonium organic salt origin DESs, 7 menthol DESs, and 17 thymol-based DESs.