Objective: This study aimed to develop an augmented reality (AR)-based system for spine surgical training that enables precise alignment of physical and virtual models and provides real-time feedback for depth perception using Unity and the Vuforia Engine. Methods: The system uses Vuforia’s Model Target to track and superimpose a virtual spine model onto a 3-dimensional-printed spine model without the need for external markers. The alignment accuracy was measured using positional and rotational error calculations via Unity’s built-in functions. The tool tracking was performed for the pedicle screw driver using ArUco markers. Results: The system achieved precise superimposition with minimal positional and rotational misalignment errors. Unity's tracking and real-time feedback ensured robust performance of the system. The system also achieved real-time tool tracking and superimposition of the real and virtual models of the tool. The system successfully demonstrated its ability to maintain spatial accuracy and provide immersive visual guidance, critical for spine surgical training. Conclusion The developed AR-based system presents a cost-effective and markerless solution for enhancing depth perception and alignment accuracy in spinal surgical education. It is suitable for use in resource-limited settings due to its reliance on widely accessible hardware and software, with potential future enhancements including real patient data and automated feedback mechanisms.