Small scale production of H2 via sorption enhanced auto-thermal reforming (SEATR) of methane is simulated using Ni based catalyst and CaO sorbent for the capturing of CO2. One dimensional heterogeneous reactor model was developed using gPROMS model builder to study the performance of SEATR reactor. At low pressure mode, the process was evaluated for varying temperature, pressure, gas flux and steam to carbon ratio. Chemical equilibrium with application (CEA), an equilibrium based software was employed so as to compare both equilibrium and simulation results. Under a range of temperature (500–1000 K), pressure (1–10 bar), S/C ratio (1–6), and O/C ratio (0.2–0.6) close to equilibrium conditions, model outputs satisfactory results with regard to CH4 conversion, CO2 capturing, H2 yield and purity. At 750 K, 2.9 bar, Gs of 0.4 kg/m2 s, S/C of 3 and O/C of 0.45, H2 purity and CH4 conversion achieved was 97% and 94% respectively in comparison with 66% and 77% from conventional auto-thermal reforming. In Bayesian Regularization (BR), Mean square error(MSE) and R value is minimum for neural network algorithm comparison. It accounts for 1.2e−10 and 0.999 respectively. BR produces minimum error with increase in Epochs and gradients values highlighting maximum performance with optimize computation time for process modeling data-integration studies and generalization.