A malaria vaccine that blocks parasite transmission from human to mosquito would be a powerful method of disrupting the parasite lifecycle and reducing the incidence of disease in humans. Pfs48/45 is a promising antigen in development as a transmission blocking vaccine (TBV) against the deadliest malaria parasite Plasmodium falciparum. The third domain of Pfs48/45 (D3) is an established TBV candidate, but production challenges have hampered development. For example, to date, a non-native N-glycan is required to stabilize the domain when produced in eukaryotic systems. Here, we implement a SPEEDesign computational design and in vitro screening pipeline that retains the potent transmission blocking epitope in Pfs48/45 while creating a stabilized non-glycosylated Pfs48/45 D3 antigen with improved characteristics for vaccine manufacture. This antigen can be genetically fused to a self-assembling single-component nanoparticle, resulting in a vaccine that elicits potent transmission-reducing activity in rodents at low doses. The enhanced Pfs48/45 antigen enables many new and powerful approaches to TBV development, and this antigen design method can be broadly applied towards the design of other vaccine antigens and therapeutics without interfering glycans.