(Disclaimer: the following might fit better on Worldbuilding - on the one hand, I'm not looking to write a story, but on the other, I don't know enough physics to know whether this is a trivial "no, save it for science fiction". Also could be considered a follow-up to this question).
Suppose for a moment that gravity is symmetric with electromagnetism, only with like "charges" attracting rather than repelling, and opposite charges repelling rather than attracting.
At the macro scale, billions of years of opposite gravitational polarities mean a sort of segregation of positively and negatively "charged" particles, which seems a plausible surface explanation for the lack of antigravity in day-to-day life - but, as I said, I don't know the physics well enough (read: at all) to say how well that explanation holds up under closer scrutiny.
At the micro scale, I don't know that we've had the capacity to observe the effects of gravity one way or the other on individual particles (to the extent that "individual particles" makes sense in modern physics), but while I'm on this hypothetical/crackpot bent I'd note that as long as the repulsive force from antigravity is exceeded by an attractive force from one of the other fundamental forces you could have, say, (obviously less stable) "hybrid" atoms, with a corresponding reduction in the observed gravitational force due to that atom.
Averaged over lots* of atoms, and you'd have what appears to be gravity that's significantly weaker than other forces, due to some of it being cancelled out by the mixed-in antigravity particles (presumably both fall off at $1/r^2$, and I'm also assuming there's no catastrophic mutual annihilation as with antimatter).
So. "totally uncharged" is obviously a consistent option. Are there any nonzero magnitudes for charged gravity that would be consistent with observations?
Probably more importantly for physics, but less important from a personal-interest perspective, how would we test this? Or, if this has already been disproved elsewhere, what experiments have we performed to rule this out?
*For experimentally confirmed values of "lots", ideally