“Whole word” is indexed and fast. “Anywhere in word” scans the whole corpus — slower, but it finds a Hebrew root behind an attached prefix.
Enter a search term above.