Advances in additive manufacturing have made truss networks a popular basis for the design of new architected materials with controllable effective properties on the macroscale, which are the result of a carefully selecting small-scale truss architecture (including the truss topology and its geometric details). The resulting effective properties can be tuned to achieve lightweight (meta-)materials with, e.g., high stiffness, strength, and toughness as well as energy absorption and directional wave guiding. The new fabrication opportunities have come with new modeling challenges, since the optimization and fine-tuning of the effective truss behavior requires theoretical-computational tools to efficiently and accurately predict the effective performance. We here discuss several such modeling strategies that have been developed in recent years : from linear and nonlinear homogenization for the extraction of effective mechanical behavior to quasicontinuum techniques for coarse-graining the discrete network, to predicting wave propagation and energy absorption in trusses beyond band gap engineering. Overall, our goal is to reduce the complexity of the discrete truss problem by appropriate model assumptions and by passing to a continuum limit where possible, rather than using brute-force calculations.