The number of pods you can have on your nodes varies based on the capacity of the host server. When you are choosing your node type think about how many pods you want to run per node/machine. The bigger the machine the more nodes you can run, however the bigger the machine the more it costs. Also for every node there are a several pods allocated for the Kubernetes system. So for example you may think to save money you should choose a t1.micro to keep costs low, however you can only really effectively have 1 worker node because a t1.micro allows for 4 pods, but 3 of them are used up by Kubernetes to manage it. So perhaps choose a t3.medium which can house 17 pods and so effectively you can have 14 worker pods.
Here is a GIT repository that lists the number of pods per machine type. This is a handy way to figure out what type of machine you want to use.