I have now access to gpt4 turbo and dalle-3 :D I had to update the python client and change a bit the response recolection as now it returns objects, but the asistan is working againa I was able to generate my first image with the imagine action \o/ The image in cuestion was not great because I was testing with a dummy prompt:
Something a bit more fancy after a couple tries (copyright protection is a bummer):
An another one: