For a continuous variable the gradient (or slope) of a cdf plot is equal to the probability density at that value. That means that the steeper the slope of a cdf the higher a density or histogram plot would look at that point:
The disadvantage of a cdf is that one cannot readily determine the central location or shape of the distribution. We cannot easily recognize common distributions like a triangular, normal, and uniform. Looking at the plots below, you will readily identify the distribution form from the left panels, but not so easily from the right panels:
For a discrete distribution, the cdf increases in steps equal to the probability of the x-value occurring:
