Watching the stars over several hours (either by making a note by eye from time to time, or using time-lapse photograph) gives convincing evidence of relative motion between stars and Earth. The fact that the stars return in their same pattern each night convinces us that the relative motion is a rotation.
To resolve the question: which is rotating, the Earth or the set of stars? general relativity offers us two interesting things to say. First, there is no need to insist on an answer which says that one perspective is 'wrong' and the other 'right'. Rather, one can set up physics either way. One can adopt a frame (A) where Earth does not rotate relative to the frame, or a frame (B) one where the stars do not move relative to the frame. Calculations of the type employed in general relativity work either way and agree with observations. However there is a difference: frame (B) can be an inertial frame, but frame (A) is not.
Let's now take the standard perspective for this kind of question, where we want to use an inertial frame. The question asks for evidence to show that Earth is rotating relative to any such local inertial frame. I have always found the Foucault pendulum to be particularly impressive here, but of course the question is right to say this cannot be accommodated on a desktop. The trouble is that anything that could sit on a desktop (e.g. a laser gyro) would have to have some quite sophisticated internal structure so someone looking at it would have to take it on trust that it is functioning how its manufacturer says, or else they would have to learn a lot of physics to understand it for themselves, or they would have to decide which group of people to trust concerning the evidence of the gyro. So with this in mind it is hard to offer really convincing evidence confined to any small device. The only way to become convinced is to widen ones perspective, so that it includes trips to the nearest science museum housing a Foucault pendulum for example.