Directed random graph with power-law degree distributions without selfloops or multiple edges

I am trying to generate a random graph with power-law degree distribution in in- and out- degree with some constrains:

1. in- and out-degree distributions should be correlated (i.e. a node with high in-degree should also have high out-degree)
2. self-loops are not allowed
3. multiple edges (in the same direction) between the same pair of nodes are not allowed
4. I need control over the number of nodes and the number of edges (or equivalently the mean degree)
5. it should be reasonably fast as I need to generate 1000 graphs with 1000 nodes and edges in the range of 1000 to 10,000 each.

I am aware of the model by Balloás but this allows self-loops and multi-edges. Implementation will be in Python. I am aware of Networkx, however none of the generators seem to fit my needs.

Thank you for your help!