Objective: Ectopic pregnancy (EP), a life-threatening condition in early pregnancy, requires accurate and timely management to reduce morbidity and mortality. While methotrexate therapy is a widely used non-invasive treatment, predicting its success remains challenging. This study introduces PROMETHEUS (Prediction of Methotrexate for Ectopic Pregnancy Treatment Success), an AI-driven decision support tool designed to assist clinicians in selecting optimal treatment strategies, aiming to minimize unnecessary surgeries and enhance patient outcomes. Methods: A retrospective nested cohort study analyzed 602 cases of EP diagnosed at Tepecik Education and Research Hospital between 2013 and 2023. Patients were grouped based on treatment received: single-dose methotrexate, two-dose methotrexate, surgery, or surgery following failed methotrexate therapy. Twenty-four clinical parameters were recorded, with 22 included in the final analysis. PROMETHEUS was developed using 15 AI algorithms, including Bagging, J48, and Random Forest. The dataset was divided into training (80%) and validation (20%) cohorts. Results: Key findings highlighted significant clinical differences between treatment groups, such as larger ectopic mass sizes and altered blood counts in patients requiring surgery after methotrexate failure. All algorithms used for modelling Prometheus achieved high accuracy (p<0.001), of them Bagging algorithm demonstrated an optimal balance between sensitivity (92.7%) and specificity (94.8%), along with the highest AUC (93.8%) (p<0.001). Conclusion: PROMETHEUS represents an innovation in ectopic pregnancy management, offering a high-accuracy AI-based decision support tool to optimize methotrexate therapy selection. This innovation has the potential to reduce treatment-related morbidity, streamline clinical workflows, and enhance personalized care. Future multi-institutional studies are needed to validate PROMETHEUS or similar AI-based applications and expand their clinical applicability.