Nanofiltration (NF) membranes have been employed in pretreatment unit operations in desalination processes. In order to predict NF performance, a systematic study on the filtration performance of NF membranes against seawater is required. In this work, NF membranes have been studied in details to check their filtration performance with salt mixtures at salinity levels similar to that of seawater. The membranes were characterized using Atomic Force Microscopy. The Spigler-Kedem model (SKM), artificial neural network (ANN), and modified Donnan Steric Pore Model (DSPM) were used to analyse the experimental data. The SKM model was used to fit the experimental data of rejection with the permeate flux in order to determine the reflection coefficient and the solute permeability. The ANN model was used to predict the rejection versus pressure and the rejection versus permeate flux. The DSPM model was modified by the inclusion of the dielectric constant term and the effect of osmotic pressure for solutions with higher concentration compared to those reported previously. Finally, the Verbene cost model was employed to check the performance and economics of NF for desalination applications.