In A Priori Justification, Casullo presents a minimal conception of the a priori as simply nonexperiential justification. This, in turn, is explained negatively as justification that does not arise due to the operation of the five senses. This clearly leaves open the possibility that there are several different kinds of a priori justification, but nearly all advocates for the a priori that I can find seem to assume that there must be a single, unified source.
A notable exception is Pap's 1944 article (Phil. Rev. 53: 465-484), although he weakens the plausibility of his three-fold distinction by concluding that the different kinds of a priori can easily intermingle.
Why not take a harder line and insist that there are different kinds of nonexperiential justification? One thought would be that some a priori justification is absolute because it is tied to conditions on concept possession, as with Peacocke, while some a priori justification is relative because it is tied to constitutive frameworks, as with Michael Friedman. Maybe this is the best way for the defender of the a priori to take on the radical empiricist.