PyTextCat[1] guesses the language of a given input text from over 70 different languages. It is an implementation of the classification technique described by William B. Cavnar & John M. Trenkle (1994) in N-Gram-Based Text Categorization[2], and is based upon Gertjan van Noord's Perl implementation[3]. textcat.py provides a command-line interface to the library. Run it with no arguments to see usage information. PyTextCat is released under the LGPLv2.1 (see COPYING.LESSER and COPYING). The lm files and test texts are from TextCat[3], and are licensed under the same license. [1] http://da.vidr.cc/projects/pytextcat/ [2] http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.53.9367 [3] http://www.let.rug.nl/vannoord/TextCat/