Single-pass tangential flow filtration (SPTFF) is a key operation in biopharmaceutical downstream processing for in-process volume reduction, and for achieving a higher final concentration than conventional ultrafiltration processes. The performance of SPTFF also impacts the overall process yield. A model-based multi-objective optimization framework is proposed for determining the optimum configuration of membranes connected in series or parallel inside a multistage SPTFF module for maximising the volume concentration factor (VCF) and the throughput. The resistance-in-series model is used for predicting the permeate flux. Single objective optimization was performed to determine maximum achievable VCF and throughput over wide ranges of feed flow rate and transmembrane pressure using the Particle Swarm Optimization. A non-dominated sorting-based genetic algorithm was used for multi-objective optimization to determine the optimum operating parameters for maximum throughput as well as maximum VCF. The optimization results were validated with published experimental data. A mixed variable multi-objective optimization framework was applied to obtain the optimal membrane configuration by minimizing the number of membranes while maximizing VCF and throughput; and for a given configuration of SPTFF, a set of optimal operating conditions were identified. These optimization results provide useful guidelines and insights for designers and plant operators to achieve cost-effective and efficient ultrafiltration.