This study describes a novel application of artificial intelligent-based computing paradigm via knacks of neural networks backpropagated through Levenberg–Marquardt scheme (NNs-BLMS) to investigate the mathematical model of Carreau nanofluid thin film flow over a stretching sheet under the magnetic field effect (CNTFFSSM). The system of PDEs of the designed model is transformed by using suitable transformations to the system of ODEs. The Adams deterministic numerical technique is utilized for generation of a dataset for the proposed technique for six varients to produce corresponding six scenarios of the designed model by varying magnetic parameter, Brownian motion parameter, unsteadiness parameter, Weissenberg number, thermophoresis number and Lewis number. The computational intelligent solver NNs-BLMS is applied to the CNTFFSSM model by performing processes on training, testing and validation samples. The efficiency of the proposed technique is validated by comparison of the standard solution and outcomes of the proposed solver for designed model through histograms, error analysis and regression analysis. The outcomes disclosed that velocity field is a decreasing function of magnetic parameter, while an increasing function of the Weissenberg number. The temperature and concentration profiles enhance with higher thermophoresis parameter, while diminish with higher Weissenberg number. Moreover, higher Lewis number causes lower concentration profile.