The percentage of radiation that is thermal versus non-thermal varies from quasar to quasar, and for a single quasar is going to be different at different wavelengths, so there isn't one simple answer (and also it depends on what angle you are viewing the AGN from relative to the jet axis)
And similarly, there isn't a single value you can quote for the accretion disk temperature. Different AGN have different black hole masses, different matter infall rates and generally different environments. And a single accretion disk doesn't have a single temperature either: the temperature is higher towards the center, but to the details of what the emission spectrum looks like depends a lot on what parameters you put in to the model and what assumptions you make about energy transfer within the disk etc.
Assuming a black body spectrum peaking in medium to hard x-rays is probably hand wavingly in the right ball park.
The majority of the emission at lower wavelengths (and even in x-rays for more extreme AGN) comes from relativistic enhanced synchrotron radiation. It is possible to estimate the fraction of synchrotron to thermal radiation from polarisation studies (synchrotron radiation is polarised depending on magnetic field strength, while thermal radiation is unpolarised). But the numbers you get are going to be somewhat model dependent, rather than objective facts.
Gamma ray and higher energy emission (yeah, they are all gamma rays, but MeV, GeV and TeV gamma rays are all different beasts) are mosty likely produced by inverse Compton scattering by electrons (either of some external photon source such at the accretion disk, or the synchrotron self Compton mechanism), although other mechanisms such as proton synchrotron mount also be plausible.
Basically, there isn't a simple, single answer to your question. Modelling the emission spectrum of AGN over 20 orders of magnitude of the electromagnetic spectrum is more involved than that, is model dependent, and varies massively between different objects.
