This study presents two predictive models, grounded in the truss analogy, to estimate the increase in the shear resistance of reinforced concrete (RC) beams strengthened with carbon fibre reinforced polymer (CFRP) bars or laminates applied with the near-surface mounted (NSM) technique. A dataset of 195 RC beams from the literature was collected, with 124 beams carefully filtered and refined for analysis. The proposed models require predictions for the effective strain in CFRP reinforcements and the inclination angle of the critical diagonal crack, which were developed in this work. These prediction models incorporate multiple influential features of the strengthened beams to enhance accuracy. Validation against the compiled dataset demonstrated a substantial reduction in mean absolute percentage error (MAPE) for predicting CFRP reinforcement contribution to shear resistance from 69% and 40% (using models by Mofidi et al. and Bianco et al., respectively) to 25% in beams strengthened with CFRP bars, and from 44%, 43%, and 28% to 21% in beams strengthened with CFRP laminates (using models by Mofidi et al., Bianco et al. and Dias and Barros, respectively). Furthermore, integrating various predictors ensured consistent error distribution across the dataset, reducing variability and increasing reliability. The reduced variability and increased reliability exhibited by these models suggest promising prospects for the future development of design models for this strengthening technique.