Biliary tract cancer (BTC), including gallbladder cancer and cholangiocarcinoma, is characterized by difficulty in early diagnosis, strong invasiveness, high lymph node metastasis rate and poor prognosis. Radical resection is an effective treatment for BTC, whereas high postoperative recurrence rate severely shortens the long-term survival of BTC patients. Prognostic models can improve the levels of precision diagnosis and treatment through prognostic evaluation and precise identification of populations with high recurrence risk. These models have important guiding values for clinical treatment decision-making such as surgery, neoadjuvant therapy, adjuvant therapy and follow-up monitoring, which contribute to elevating the quality of life and prolonging the survival of BTC patients.With recent developments of interdisciplinary integration, on the basis of traditional methods, prognostic models for BTC based on machine learning, radiomics and other technologies have been continuously developed, showcasing certain advantages. In this article, development status and challenges in prognostic models for BTC were illustrated, aiming to provide reference for clinical practice and research in this field.