The Lagrangian $$\mathcal L = -\frac{1}{4}F_{\mu\nu}F^{\mu\nu} + \mathcal L_\textrm{free} + eA_\mu J^\mu \tag{1}$$
where $A_\mu$ is the 4-potential, $F_{\mu\nu} = \partial_{[\nu}A_{\mu]}$ is the field tensor, $\mathcal L_\textrm{free}$ describes fields other than $A_\mu$, and $J^\mu$ is the 4-current density expressed in these other fields, describes a QED-like theory. When $\mathcal L_\textrm{free}$ describes a free Dirac field $\psi$ and $J^\mu = \overline\psi\gamma^\mu \psi$, it is precisely QED. The Dirac field has spin $\frac 1 2$
We can instead take $$\mathcal L_\textrm{free} = \frac{1}{2}\big( (\partial_\mu a)(\partial^\mu a^\dagger) + m^2 aa^\dagger)$$
with $$J^\mu = i(a\partial^\mu a^\dagger - (\partial^\mu a)a^\dagger). \tag{2}$$
The field $a$ describes spin $0$ particles.
The theory described by (1) is as self-consistent as QED is, that is, it is renormalizable. This is because the necessary and sufficient ingredient in the renormalizability of QED is that the constant $e$ is dimens.ionless (in natural units). With $J^\mu$ according to (2), this is the case.