This study investigates the radiative magnetohydrodynamic (MHD) flow of a chemically reactive tangent hyperbolic nanofluid over an exponentially stretching sheet under Prescribed Exponential Surface Temperature (PEST) and Prescribed Exponential Heat-Flux (PEHF) conditions. The problem is important due to its applications in thermal management, polymer processing, cooling technologies, and advanced heat and mass transfer systems. Using suitable similarity transformations, the governing nonlinear partial differential equations are reduced to coupled ordinary differential equations and solved numerically using MATLAB’s bvp4c solver. To enhance predictive capability, a feed-forward Artificial Neural Network (ANN) with architecture [5–10–1] is developed to predict the local Nusselt and Sherwood numbers from the numerical dataset. The results reveal that thermal radiation significantly enhances the temperature distribution and thickens the thermal boundary layer, whereas the chemical reaction parameter decreases nanoparticle concentration and reduces the solutal boundary-layer thickness. Comparative analysis further shows that the PEHF case produces stronger thermal and concentration distributions, while the PEST case provides slightly better ANN prediction accuracy and numerical stability. The ANN models exhibit excellent agreement with the bvp4c solutions, achieving extremely low mean squared errors in the range of \(10^{-10}\) – \(10^{-5}\) with regression coefficients close to unity \((R \approx 1)\) . The proposed computational model provides a reliable, accurate, and computationally efficient approach for predicting coupled heat and mass transfer behavior in complex non-Newtonian nanofluid systems.
