Low metallicity stellar populations are very abundant in the Universe, either as the remnants of the past history of the Milky Way or similar spiral galaxies, or the young low metallicity stellar populations that are being observed in the local dwarf galaxies or in the high-z objects with low metal content recently found with JWST. Our goal is to develop new high-spectral-resolution models tailored for low-metallicity environments and apply them to analyze stellar population data, particularly in cases where a significant portion of the stellar content exhibits low metallicity. We used the state-of-the-art stellar population synthesis code HR-pyPopStar with available stellar libraries to create a new set of models focused on low metallicity stellar populations. We have compared the new spectral energy distributions with the previous models of HR-pyPopStar for solar metallicity. Once we verified that the spectra, except for the oldest ages that show some differences in the molecular bands of the TiO and G band, are similar, we reanalyzed the high resolution data from the globular cluster M 15 by finding a better estimate of its age and metallicity. Finally, we analyzed a subsample of mostly star-forming dwarf galaxies from the MaNGA survey we found similar stellar mass-mean stellar metallicity weighted by light to other studies that studied star forming dwarf galaxies and slightly higher mean stellar metallicity than the other works that analysed all types of dwarf galaxies at the same time, but are within error bars.