Pack architecture emerges as a critical design parameter in battery electric vehicles, since it establishes the dominant fraction of the cost, usable range, pack weight, thermal load, and servicing value of the system. The present work develops a discrete multi-objective model for selecting pack architectures in cell-to-pack BEV systems, focusing on the evaluation of range, battery cost, and degradation cost per kilometre at the vehicle level. The candidate pool consists of 864 different pack architectures, created from the combination of two technology generations (2025 and 2030), three cell geometries, four gross energy classes, four levels of structural integration, three fast-charging rates, and three cycle lives. Cell-level density parameters are transformed into pack mass and volume, gross energy into usable energy, range is determined based on a mass-dependent consumption correlation, and lifetime is quantified in service-equivalent distance travelled at the end of life. Filtering candidates based on charging time, pack mass and volume limits leaves 747 candidate packs, while excluding redundant charge rates reduces the number of feasible architecture-lifetime combinations to 249. These results address the fundamental design question head-on by indicating that the largest battery does not yield optimal system performance in the representative C-segment crossover application. The optimum compromise design of 2025 technology consists of a 65 kWh prismatic pack at ψ = 1.15 with 2500 cycles, while the 2030 optimum has a capacity of 75 kWh with identical architecture parameters. High levels of integration and long cycle life consistently expand the achievable trade space, and large packs provide additional range at the cost of increased mass and associated higher costs.