You can freely combine Gluon with standard MXNet and NumPy modules—for example, module, autograd, and ndarray —as well as with Python control flows.

MXNet supports Python, R, Scala, Julia, and C++, but its best-supported API is for Python. MXNet has demonstrated good scaling (85 percent of linear) across multiple GPUs on multiple hosts.

Choosing MXNet might come as a surprise to some, given the number of other frameworks — TensorFlow, Theano, Torch, or Caffe, to name a few — with far broader name recognition.