This study investigates the efficacy of sequence modelling in predicting student performance. Over an eight-month period, user interactions were recorded as students solved navigation tasks in the SystemX programming environment. On this data, three sequence models were trained to learn the temporal dependencies for several performance features. We compared the models’ predictive capabilities and found the Transformer architecture to perform best in making multi-step predictions. Although prediction quality declines for multi-step forecasts due to the accumulation of error, we show that the quality of long-term forecast become closer to those of short-term forecasts, as the input length increases. Our results provide valuable insights for the development of more effective teaching tools that can monitor and support student learning in real time.