Selectively and sensitively detecting specific exosomal markers is critical for early diagnosis of liver cancer. However, identifying specific exosomal biomarkers and establishing accurate, convenient detection methods remain challenging. In this study, we used bioinformatics to identify the higher levels of EpCAM and GPC-3 proteins on liver cancer exosomes. These markers were used to create a dual-antibody functionalized transistor biosensor for precise detection of liver cancer exosomes. The techniques exhibited outstanding specificity and sensitivity. Detection thresholds in PBS and simulated plasma were established at 20 particles/μL and 47 particles/μL, respectively, facilitating the distinction of liver cancer cell-derived exosomes from those originating from various other cancer cells. Furthermore, in clinical samples testing, this approach not only distinguished clinical samples among liver cancer patients and healthy individuals, but also demonstrated the ability to differentiate liver cancer from other types of tumors, achieving a precision and accuracy rate of 100 %. The developed biosensor demonstrates excellent potential for clinical application and this work offers a promising and effective approach for cancer diagnosis.