That's not true. He had perfectly good reasons for atomism in his context.

The ontological arguments of Parmenides (and as exposited by Melissus) lead to extremely unpalatable, if not outright contradictory, conclusions, such as there being no time or change or different entities. The arguments seem valid, and most of their premises are reasonable, but one of his most important and questionable premises is that void cannot exist.

Reject that premise and you are left with matter and void. How are matter and void distributed? Well, either matter can be indefinitely chopped up (continuous) or it must halt and be discrete at some point. The Pluralists like Anaxagoras take the former approach, but continuousness leads to its own issues with regard to change.* So to avoid issues with infinity, you must have discrete matter with size/divison limits - atoms.

So, Democritus and Leucippus are led to Atomism as the one safe path through a thicket of paradoxes and problems. Describing it as wild conjecture is deeply unfair, and, I hope, ignorant.

* One argument, if I remember it from Sextus Empiricus's Against the Physicists correctly, is that if matter really is infinitely divisible, then you should be able to divide it again and again, with void composing ever more of the original mass you started with; if you do division infinitely, then you must end up with nothing at all! That is a problem. Cantor dust would not have been acceptable to the ancient Greeks.