ABSTRACT: I briefly review observations of the radio spectrum of Sgr A*, its variability, and the evidence for a submm-bump. The latter, if real, indicates an extremely compact size of Sgr A* which in principle would allow to directly image the suspected supermassive black hole with future mm-VLBI experiments. For this, however, we need to understand the nature of this source which is currently hotly debated. I show that a jet model can reproduce well all the observed radio properties of Sgr A*, i.e. the spectrum and the size (limits) as a function of frequency. The model is very simple and has extremely few free parameters but nevertheless can be used to fit other jet sources like NGC 4258, M81*, or even GRS 1915+105 without twisting the parameters. However, while in these other sources the (derived) jet power always seems to be comparable to the associated accretion disk luminosity, the NIR upper limits of Sgr A* severly constrain the structure of any associated accretion disk--if one ignores the possibility of intrinsic obscuration. The jet model indicates that the accretion rate near the black hole has to be at least ~10-7Mo/yr while the NIR fluxes allow only much lower accretion rates if a standard alpha -disk is assumed. Therefore the jet model supports a non-standard accretion disk in Sgr A*, which could either be described by an ADAF or an EDAF model (an Ejection Dominated Accretion Flow, where most of the locally produced energy is channeled into an outflowing wind.)