The development of accurate models to predict the solvation, solubility, and partitioning of nonpolar and amphiphilic compounds in aqueous environments remains an important challenge. We develop state-of-the-art group-interaction models that deliver an accurate description of the thermodynamic properties of alkanes and alcohols in aqueous solution. The group-contribution formulation of the statistical associating fluid theory based on potentials with a variable Mie form (SAFT-gamma Mie) is shown to provide accurate predictions of the phase equilibria, including liquid-liquid equilibria, solubility, free energies of solvation, and other infinite-dilution properties. The transferability of the model is further exemplified with predictions of octanol-water partitioning and solubility for a range of organic and pharmaceutically relevant compounds. Our SAFT-gamma Mie platform is reliable for the prediction of challenging properties such as mutual solubilities of water and organic compounds which can span over 10 orders of magnitude, while remaining generic in its applicability to a wide range of compounds and thermodynamic conditions. Our work sheds light on contradictory findings related to alkane-water solubility data and the suitability of models that do not account explicitly for polarity.