Loss-of-function mutations in the gene encoding progranulin (PGRN) are a common cause of frontotemporal dementia, leading to markedly decreased PGRN levels. A potential therapeutic strategy is therefore to increase extracellular PGRN by blocking sortilin-mediated PGRN clearance. Here, we describe the systematic design and optimization of small biparatopic sortilin-binding proteins based on the non-immunoglobulin affibody scaffold. Two anti-sortilin affibody molecules were genetically fused into a panel of heterodimeric constructs exploring multiple domain orientations, linker lengths, and helix truncations. In total, ten distinct dimer variants were generated and evaluated for binding and functional activity. Optimization of domain arrangement and truncations yielded constructs exhibiting subnanomolar sortilin affinities, corresponding to up to an approximately 45-fold improvement over the strongest monomeric affibody and pronounced avidity effects. In a PGRN clearance assay, the lead 18.6-kDa dimer efficiently increased extracellular PGRN levels with an EC50 value of 0.32 nM and produced substantially greater PGRN fold changes than monomeric constructs. This PGRN elevation was accompanied by marked reductions in both cell surface and total sortilin levels, consistent with effective receptor blockade and modulation. Together, these results demonstrate how systematic combination and geometric optimization of two individually moderate-affinity binders can generate a highly potent biparatopic inhibitor, as well as illustrating the versatility of compact and modular affibody molecules as building blocks in therapeutic protein design. Given the growing amount of evidence implicating sortilin in neurodegeneration, inflammation, and cancer, biparatopic affibody-based inhibitors may enable exploration of this pathway in diverse biological contexts, both as research tools and as starting points for drug development.