At this time, it supports only English.
If you give it a text in a different language, it may still give you a result based on some sentence structure metrics, but the result is meaningless.