The vSphere hosts in the environment handle encapsulation/decapsulation of the VXLAN so your Nexus switches don't need to handle them any differently than other IP traffic. The only potential change needed is that your Nexus switches will need at least a 1600 byte MTU to handle the encapsulation overhead.
Encap and De-cap is not been done by the Nexus, so there is no vxlan config needed on Nexus.
In NSX, encap and Decap will be done at the host level (By the vtep created on the host).
You just need to configure jumbo mtu on north-bound switches.