Strain hardening properties of AA5182-O metal sheet are experimentally studied and analytically modeled at different temperature in this research using a dogbone specimen. The strain rate hardening and thermal softening behaviors are then modeled by popular analytical models to consider the strain rate and thermal effects. The high nonlinear strain hardening with the effect of strain rate and temperature is also illustrated by artificial neural network. The results show that the strain hardening with strain rate and thermal effect can be accurately modeled by the neural network compared with the conventional models, but its computation efficiency is much lower than the analytical models.