Are 'Listen' and 'Silent' anagrams?
Yes, and they are the default example. With ignore case on, both normalise to the same six letters, e-i-l-n-s-t, so every letter count matches. That pair is the textbook example precisely because it works with no punctuation or spacing tricks.
Do spaces and punctuation count?
Not by default. The ignore-spaces and ignore-punctuation options are on, which is what lets 'Dormitory' match 'Dirty Room'. Switch them off to compare the raw characters, which is the strict definition and treats a space as a character that has to appear the same number of times in both phrases.
Why are café and face not anagrams here?
Because é and e are different Unicode characters, and the tool normalises to NFC without folding accents. This is the strictly correct answer for text processing. If you want them to match, strip the diacritics first, for example by normalising to NFD and removing the combining marks.
Does it find anagrams of a word for me?
No. It checks a pair you supply. Generating anagrams is a different problem: it needs a dictionary and a search over permutations, and it produces a list rather than a yes or no. This tool answers the verification question.
What is a perfect anagram versus an imperfect one?
A perfect anagram uses every letter of the source exactly once, which is what this tool checks. An imperfect anagram, sometimes called an antigram or a partial anagram, drops or adds letters. If the verdict comes back negative, the letters-that-differ panel shows you exactly which letters are out of balance and by how much.
Is the comparison case sensitive?
Only if you turn ignore case off. With it on, everything is lowercased before comparison. Off, an uppercase L and a lowercase l are counted separately, so 'Listen' and 'silent' would fail on the L alone.