Copy some Latin text into the box, leaving spacing, punctuation, and capitalization in any format you wish. [clear form]
I use this as a first pass when annotating large texts* for vowel length. I then adjust its output by hand, which is easier than typing every macron from scratch, since most macrons are easily predicted. It should get about 90-95% of vowels correct (drawing on a 100,000-word annotated corpus and phonological heuristics). It doesn't do any kind of parsing, so it has no way of knowing whether an instance of, say, femina represents fe-mina or fe-mina-. (It will always guess the former, which is the more frequent.) *If you're annotating only a few words or lines, skip this and go straight to a dictionary (e.g. Google each unfamiliar form and click the wiktionary.org result).