$\partial_t f(x,t)= \alpha \partial_x^2f+\beta f + \gamma g \\ \partial_t g(x,t)= \alpha \partial_x^2g -\beta f - \gamma g$
With everything real.
I tried to take the first equation and express g(x,t) as function of f(x,t)
$g=\frac{1}{\gamma}\partial_t f-\alpha \partial_x^2f-\beta f$
and substitute g in the second expression. Then trying to solve with Fourier transform doesn't seem to work.
Or adding both equation leads to
$\partial_t(f+g)=\alpha\partial_x^2(f+g)$.
Assuming variables can be separated for (f+g) this can be solved for (f+g) like a wave propagation equation. But anyway this does not solve the problem.
Thanks for indications.

Introduce the new unknown functions $$u:=\beta f+\gamma g,\qquad s:=f+g\ .$$ Then the system gets decoupled: $$u_t=\alpha u_{xx}+(\beta-\gamma) u,\qquad s_t=\alpha s_{xx}\ .$$ Maybe there will be difficulties when $\beta=\gamma$; see Juan Ospina's answer.