Immune checkpoint inhibitors (ICIs) targeting PD-L1 and CD47 are clinically limited by severe off-target toxicities. To address this issue, immunotherapeutic prodrug strategies have been developed, aimed at preventing antibodies from binding to targets in healthy tissues and thereby reducing systemic toxicity. Existing strategies include prodrug technologies that mask the active sites of antibodies via peptide or polyethylene glycol (PEG) modification—yet these approaches also cause antibodies to lose their targeting ability. Herein, we propose an antibody prodrug strategy (termed FA-PEG-S-Ab) with active targeting capability. By modifying antibodies with folate-PEG-disulfide and PEG-disulfide linkages, we developed two novel prodrugs: FA-PEG-S-Atz (PD-L1-blocking prodrug) and FA-PEG-S-Hu5 (CD47-blocking prodrug). This strategy functions through two key steps: first, folate binding to folate receptor α (FRα)-mediated tumor-specific targeting enables the prodrugs to accumulate specifically in tumor tissues; subsequently, the high concentration of glutathione (GSH) in the tumor microenvironment (TME) specifically cleaves the disulfide bonds, removing the PEG shield, releasing the antibody, and restoring the antibody’s antigen-binding activity. In vitro experiments confirmed that the modified antibody prodrug FA-PEG-S-Hu5 exhibits high affinity for FRα (K_D = 4.02 × 10⁻⁹ M) and effectively masks the antibody’s binding activity (K_D from 1.05 × 10⁻¹¹ M to 2.10 × 10⁻⁸ M). Following activation by GSH in the TME, this masking effect is reversed, and the antibody regains its binding affinity (K_D = 2.14 × 10⁻¹⁰ M). Crucially, FA-PEG-S-Hu5 completely eliminates hemolytic toxicity. This “folate targeting delivery + TME activation” prodrug strategy is expected to provide a new solution for addressing the off-target toxicities of conventional ICIs.