Malignant solid tumors, characterized by their persistently high incidence and mortality rates, pose a significant threat to human health and life, imposing a substantial societal burden. Molecular imaging enables the non-invasive, in vivo visualization of tumorigenesis and progression at the molecular level. Compared to traditional morphology-based imaging techniques, molecular imaging provides more precise information for early tumor diagnosis, treatment efficacy assessment, and clinical disease management. 18F-FDG PET/CT imaging is currently the most widely used molecular imaging modality. However, under immunotherapy, FDG accumulates extensively in activated T cells, leading to increased false-positive evaluations. It fails to effectively distinguish metabolic hyperactivity between proliferative tumor cells (indicative of true progressive disease) and infiltrating immune cells (associated with pseudoprogression), thereby complicating the assessment of immunotherapy efficacy. Therefore, exploring novel molecular imaging probes with high specificity is of critical importance for patients undergoing tumor immunotherapy, as it can lead to more accurate evaluation of treatment efficacy. Granzyme B (GZMB), a serine protease released from cytoplasmic granules of cytotoxic T lymphocytes (CTLs) and natural killer (NK) cells, induces apoptosis in target cells, particularly tumor cells-a central mechanism of tumor immunotherapy. This makes GZMB a promising molecular target for evaluating immunotherapy efficacy. This study aims to assess tumor immunotherapy outcomes using GZMB-targeted PET imaging and compare its performance with 18F-FDG PET/CT. The goal is to achieve timely and accurate efficacy evaluation and longitudinal monitoring, identify potential beneficiaries, optimize clinical decision-making, and ultimately deliver personalized precision treatment to improve overall treatment outcomes.