MHC class I molecules report on the health of the cells through the presentation of cytosolic peptides to the cytotoxic T-cells. To characterize T-cell receptor recognition, we need to obtain detailed surface features of the peptide-MHC (pMHC) complexes facing T-cell receptors. Therefore, we need high-resolution structural models of pMHC complexes. In this work, we present RosettaMHC, a pMHC homology modeling approach, which utilizes structural templates derived from the PDB using MHC groove sequence identity and peptide dihedral angle similarity. We demonstrate that, our method is pan-allelic and the structural models predicted using our approach are < 1Å backbone RMSD from the native. In conclusion, we establish that, for the most frequent alleles, we have sufficient knowledge from the existing databases to predict sub-angstrom models thereby reducing the efforts required to experimentally determine the pMHC structures.