Good question! The magnitude Bode plot depicts only the relative gain between two signals. Likewise, the phase plot gives you the relative phase difference between input and output. If you provide a known system (e.g. G(s) = 1/(s+1)) for the bode function, you already know the relative gain and phase differences. For a given input, you can find the output magnitude directly.
What you are describing is referred to as swept-sine testing which is one potential method to determine the Bode plot (and identify G(s)) for a particular system. You could equally generate the Bode plot using an impulse hammer response or a white noise input. If you want to learn more, here is a link to some material that discusses and contrasts the different methods:
As far as your original question is concerned, the documentation for the bode function provides information on all possible inputs and changes you can make. You can find the documentation here:
Hope this helps clarify things up!