Stellar X-ray and extreme ultraviolet (XUV) emission drives the heating and chemical reactions in planetary atmospheres and protoplanetary disks, and therefore, a proper estimation of a stellar XUV spectrum is required for their studies. One proposed solution is to estimate stellar atmospheric heating using numerical models, although the validation was restricted to the Sun over a limited parameter range. For this study, we extended the validation of the model by testing it with the Sun and three young, nearby solar-type stars with available XUV observational data ($ Ceti, $ UMa, and EK Dra). We first tested the model with the solar observations, examining its accuracy for the activity minimum and maximum phases, its dependence on the loop length, the effect of loop length superposition, and its sensitivity to elemental abundance. We confirm that the model spectrum is mostly accurate both for the activity minimum and maximum, although the high-energy X-rays ($ \ nm $) are underestimated in the activity maximum. Applying the model to young solar-type stars, we find that it can reproduce the observed XUV spectra within a factor of 3 in the range of 1-30 nm for stars with a magnetic flux up to 100 times that of the Sun ($ Ceti and $ UMa). For a star with 300 times the solar magnetic flux (EK Dra), although the raw numerical data show a systematically lower spectrum than observed, the spectra are in good agreement once corrected for the effect of insufficient resolution in the transition region. For all young solar-type stars, high-energy X-rays ($ \ nm $) are significantly underestimated, with the deviation increasing with stellar magnetic activity. Furthermore, our model-based estimation shows performance that is comparable to or surpasses that of previous empirical approaches. We also demonstrate that the widely used fifth-order Chebyshev polynomial fitting can accurately reproduce the actual differential emission measure and XUV spectrum. Our findings indicate that the stellar XUV spectrum can be reasonably estimated through a numerical model, given that the essential input parameters (surface magnetic flux and elemental abundance) are known.