The load-displacement curve is a primary representation used to characterize the flexural behavior of concrete beams strengthened with prestressed near-surface mounted (NSM) fiber-reinforced polymers (FRPs). However, accurately predicting the load-displacement response remains challenging due to the complex failure mechanisms and the limited experimental data of strengthened beams. In this work, a simulated dataset consisting of load-displacement responses and various structural parameters of a prestressed NSM carbon FRP-strengthened beam was generated through finite element analysis. Subsequently, artificial neural network (ANN) and long short-term memory (LSTM) models, trained on this simulated dataset, were developed to directly predict the load-displacement response based on varying structural parameters of the FRP prestress, FRP bond length, concrete type, and FRP type. The ANN model demonstrated superior prediction accuracy compared to the LSTM model, achieving an R²of 0.98 and a mean absolute percentage error (MAPE) of 2.55 %. Moreover, to enhance the model’s performance in scenarios with limited datasets, a transfer learning (TL) approach was employed. In particular, a two-step TL strategy involving pre-training with a source domain followed by fine-tuning with a target domain was implemented. Results indicated that the TL-enhanced ANN model accurately predicted the load-displacement response of prestressed NSM FRP-strengthened beams using only 8 samples from the target domain. Compared to the conventional ANN model trained exclusively on the limited target domain, the TL-enhanced ANN model achieved a 39 %-52 % improvement in MAPE across three typical cases. Finally, the developed TL technique was applied to transfer knowledge from simulated data to real experimental data. An ANN load-displacement prediction model was first trained on the simulated dataset and then fine-tuned using 10 experimental samples. The TL-enhanced ANN model demonstrated strong performance in predicting the experimental load-displacement response of concrete beams strengthened with prestressed NSM carbon FRPs.