I'm dealing with an interactive work of fiction right now, and one proposal is to combine two types of puzzles with a magnetic gas. I know that fermionic $\ce{^6Li}$, when cooled below 1 K$\pu{1 K}$, has shown verifiable ferromagnetic properties, but that is far too cold to be applicable to the work's setting.
(I'm only willing to suspend the user's belief so much.)
I know that on heating above the Curie temperature for a material, magnetic properties dissolve; but what that temperature is varies quite a bit. (I've seen them between 90 K$\pu{90 K}$ and 1400 K$\pu{1400 K}$, which is pretty radical.) Has anyone yet discovered another molecule that remains magnetic in a gaseous phase? Are any of them at least hypothesized?