The IRST Language Modeling (IRSTLM) Toolkit features algorithms and data structures suitable to estimate, store, and access very large n-gram language models. Our software has been integrated into a popular open source Statistical Machine Translation decoder called Moses, and is compatible with language models created with other tools, such as the SRILM Tooolkit.

IRSTLM is released under the GNU Library or Lesser General Public License version 2.0 (LGPLv2).

IRSTLM can be downloaded from the irstlm Github repository. Together with the source code, you will get the documentation.

A suite of regression tests for IRSTLM is available in the irstlm-regression-testing Github repository.

Application cases:

over 10,000 downloads!

Contact us: cettolo[at]