Functionally, both implementations will produce the same result with the only difference being that for static methods the function is attached to the class’ prototype. If you don’t actually need that behavior (unlikely that you do), I’d keep you static methods as separate functions as this is better for tree-shaking and code-splitting (unless you have a LOT of static methods though, these will be only minor gains). In practice, I find that static methods usually only appear for two reasons: (1) whoever wrote the code is used to Java and (2) some IDEs present warnings that you should mark methods as static if they don’t use instance variables.
If you just want that ability to namespace a bunch of static methods, you can still do so by exporting an object with all of the individual functions attached.