Continuous distribution in scipy that looks like two powerlaws?

I need to generate random samples from a distribution that looks roughly like this:

To generate the image above I combined two powerlaws:

y = a * x**(a - 1) + b * x**(b - 1)

with parameters a, b controlling the shape of the distribution. Ideally a single shape parameter would be better.

There are dozens of distributions defined in scipy's Continuous distributions section but I haven't been able to find one that matches what I need.