This study aims to develop a predictive model for the thermal conductivity of graphene nanoplatelets/SAE10W oil nanofluids using artificial neural networks and response surface methodology. Generally, the property of thermal conductivity has been measured to enhance the heat transfer efficiency of traditional heat transfer fluids. Experiments were conducted using a thermal constants analyzer at different operating conditions, such as varying the volume concentrations of nanoparticles from 0.050% to 0.150% and increasing the temperature from 20?C to 80?C. Results showed an improvement in the thermal conductivity of the nanofluid, ranging from 19% to 41%. A single hidden layer with 12 neurons was found to be the most effective architecture for the artificial neural network model. Additionally, a response surface was closely fitted to experimental data points in the response surface methodology. Then, mean squared error, root mean square error, and R-squared values were employed to validate the accuracy of the predicted models. The correlation coefficients of the artificial neural network and response surface methodology models were 0.99761 and 0.9877, respectively. Also, the accuracy of the models was assessed in terms of margins of deviation. The margin of deviation for the artificial neural network model ranged between +0.3926% and -0.4640%, whereas for the response surface methodology model, it was between +0.4137% and -0.4166%. The comparison of the artificial neural network model indicates greater accuracy than the response surface methodology technique. This method for predicting the thermal conductivity of graphene nanoplatelets /SAE10W oil nanofluids is both cost-effective and inventive, minimizing experimental research durations.