I think you are mixing two things: gradient and divergence.
The gradient is (normally) used when you have a scalar field, or function. A scalar field (or function) is when you associate a number to every point is space.
The divergence is (again, normally) used when you have a vector field, or function. A vector field (or function) is when you associate a vector (let's say x,y and z) to every point in space.
The magnetic field is a vector field, so you would normally just look at the divergence of that field. That being said, the vectorial gradient is something that exists, but is at a much higher level (https://math.stackexchange.com/questions/156880/gradient-of-a-vector-field).
Now, if you meant "what is the divergence of the magnetic field?" The answer is 0, because it's always 0. This is actually an experimental truth and if you want to understand this, i'll refer you to this nice explanation:
Why is the divergence of a magnetic field equal to zero?
Thanks to @NeuroFuzzy for the pointer. So, to complement my answer, if by gradient you mean variation over a small distance of the magnetic field, then i'll compare three situations:
1. For an infinite wire (think of your power lines), the magnetic field will run around the wire following the right hand rule. The intensity of the field will decrease following ~ 1/distance. In this case, you would say that the gradient is not very high because it doesn't vary a lot with distance.
For a loop of current, the field at a distance from the center of the loop is a complicated equation. To get an intuitively pleasing result, you need to approximate how the field behaves at a good distance. This gives a field that derceases following ~ 1/distance^3, which means that it has a very strong gradient - it varies a lot.
Lastly, my trick for visualizing a magnet is to think of it as a multitude of infinitely small current loops. Admitting this, and doing some (rather tough) math, you'll eventually find a formula (see eq. 8) which, for the same approximation as in case 2, give a variation of ~ 1/distance^3 again.
So, the final answer, would be to say that the gradient of your magnet is stronger than that of a power line - it varies more!
Hope it helps!