Start with mice, because it's much easier to get ACUC approval for mice, and monkeys are expensive. Assuming your Ab is a mouse monoclonal (MAb), you'll need an antigen binding assay (probably ELISA) as readout for your PK experiment. First you'll want to optimize your assay for mouse serum. So make a standard curve using your MAb spiked into normal mouse serum. Then do PK with one mouse and just a few time points. The results will inform your design of subsequent experiments, including calculation of statitical power. You will need several experiments, eventually with multiple mice per time point, before you come up with a publishable curve.

Only do monkeys if you really expect this MAb will go into primates without further modification. Your ACUC will require extensive work in mice to consider primate work. And you'll need staff with primate experience, because bleeding them is more complicated than bleeding mice.

Generally you should not reuse animals for a second round of injection unless you are intentionally measuring their secondary immune response that rapidly clears your MAb. Monkeys probably will develop robust anti-mouse Ab after a single challenge. Mice are likely to develop some amount of anti-idiotype response, which can be more challenging to characterize.