For intrinsic (pure) material the electron and hole concentrations are equal and depend exponentially on the inverse temperature. (The holes do not show on the plot when p=n: the hole plot is overwritten by the electron plot.) The model assumes no band degeneracy in either conduction or valence bands. It does include the temperature dependence of "the number of thermally accessible states" near the edges of the bands.
Put a straight edge along the semilog plot of carrier density. Why is the plot not really straight?