Inverse one matrix in cuda

Hello, what’s the best way to inverse one matrix using cuSolver?
In openblas I usually use one call of dgetrf, then one call of dgetri.
In cusolver I found the implementation of cusolverDngetrf() but didn’t find cusolverDngetri().
On the other hand, i managed to inverse a matrix by using a batched call with a batch size of 1
using cublasDgetrfBatched and cublasDgetriBatched.
Can you please implement the missing cusolverDngetri() for consistency?
Many thanks!!

Seems you first have to calculate a matrix decomposition (LU or QR) and then solve it (possibly for each column of the identiy matrix I , when solving the matrix equation A * X = I). Note that the matrix X is the inverse of A. See gpu - Solving dense linear systems AX = B with CUDA - Stack Overflow and algorithm - QR decomposition to solve linear systems in CUDA - Stack Overflow