Ah, that does make sense. I was aware that it’s also used in most photosynthesising organisms (that’s why it’s so abundant in real life) but I figured it could also apply to most imaginable energy sources.
The actual carbon fixation IRL takes place outside the thylakoids, though in for example plants it takes place outside the thylakoids but still inside the chloroplast. One option could be to separate the carbon fixation from thylakoids altogether, and only have it as a separate organelle (and funnily enough, light-dependent organisms that can’t make their own glucose do exist). In that case, you would probably want to still have chloroplasts (and equivalents) directly produce glucose.