The fabrication of Inconel 718 (IN718) sheet metal components often requires larger deformation loads at room temperature. In this regard, deformation of the material at elevated temperature is a promising approach for reducing the forming load and enhancing the formability. Hence, the flow-stress behavior of IN718 sheets at elevated temperatures within the range of 773–973 K over wide ranges of strain rate (from 0.001 to 1 s−1) was studied by uniaxial tensile testing. The peak load reduced significantly by 75.6 and 8.5 % at 923 K and 0.001s−1 compared with room temperature and 773 K, respectively. Also, the total elongation improved by 65.4 and 16.5 % at 923 K with respect to room temperature and 773 K, respectively. In addition, a substantial improvement in the total elongation was observed with decrease in strain rate at higher temperatures. Seven different constitutive models, viz., Johnson-Cook (JC), modified-JC (m-JC), modified-Arrhenius equation (m-ARR), mechanical threshold stress (MTS), Rusinek-Klepaczko (RK), modified Zerilli-Armstrong (m-ZA), and the artificial neural network (ANN) were developed to describe the deformation behavior of IN718 sheet material at elevated temperatures and varying strain rates. Furthermore, suitability of these developed models was determined by comparing three standard statistical parameters, namely correlation coefficient (R), average absolute error (Δ), and standard deviation (SDA). The results showed that m-JC and m-ZA models predicted the flow stress very well in accordance with the experimental data. However, the trained ANN model could predict the flow-stress behavior more accurately throughout the entire testing condition. Though the ANN model was the best among all seven models, it was strongly dependent on an extremely good set of experimental data. Hence, the physical-based m-ZA model was considered to be the suitable model that could precisely predict the flow-stress behavior of IN718 sheet material.