In clinical practice, anti-vascular endothelial growth factor (VEGF) therapies have been effectively employed to treat neovascular eye diseases. However, several clinical challenges remain unresolved: approximately 20 % of patients exhibit suboptimal responses to anti-VEGF monotherapy, and the frequent administration of intravitreal injections imposes a significant burden on patients. To overcome these challenges, we developed a novel antibody fusion protein, GB10, which consists of a VEGF-Trap and an anti-angiopoietin 2 (Ang-2) variable heavy domain of heavy-chain antibody to simultaneously inhibit the pro-angiogenic pathways of VEGF and Ang-2 for enhanced and more enduring efficacy. GB10 bound VEGF and Ang-2 with high affinity, and exhibited better activity in vitro in inhibiting the VEGF and Ang-2 signaling pathways compared to faricimab, the only marketed bispecific antibody for neovascular eye diseases. In vivo , GB10 demonstrated greater efficacy and durability compared to faricimab in a non-human primate model of laser-induced choroidal neovascularization. GB10 also possessed a longer half-life in vitreous measured in a rabbit model. Furthermore, GB10 exhibited good injectability and stability at a high-concentration of 140 mg/mL. Given its demonstrated efficacy and developability profile, GB10 represents a potential therapeutic candidate for neovascular eye diseases and warrants further investigation through clinical evaluation. • A novel antibody fusion protein designed for multi-pathways inhibition. • Better in vitro and in vivo efficacy compared to the launched drug. • No notable ocular side effects were observed, suggesting potential safety. • Improved clinical compliance due to prolonged half-life and high concentration. • Demonstrated good developability for manufacturing and clinical applications.