# Install cmuclmtk with Homebrew, MacPorts, zypper

Language model tools (from CMU Sphinx). Version 0.7 via Homebrew; verified 2026-07-18. Also installable with zypper: sudo zypper install cmuclmtk.

## Install

```sh
sudo av install brew:cmuclmtk
```

Additional install commands:

### macOS

- Homebrew (100%):

```sh
brew install cmuclmtk
```

  Evidence: local Homebrew formula metadata

- MacPorts (94%):

```sh
sudo port install cmuclmtk
```

  Evidence: MacPorts ports tree: audio/cmuclmtk/Portfile from https://api.github.com/repos/macports/macports-ports/git/trees/master?recursive=1

### Linux

- zypper (92%):

```sh
sudo zypper install cmuclmtk
```

  Evidence: openSUSE Tumbleweed package metadata: cmuclmtk from https://download.opensuse.org/tumbleweed/repo/oss/repodata/50b07339cb64c8ed4091bdbabddadc1ff5737b090e478818a195b40d8a3292861a879139b4a3987c31109699fde9fbf4a716367ddf4eef77da75f96e3193d6ed-primary.xml.zst

## Package facts

- **Package key:** brew:cmuclmtk
- **Package manager:** Homebrew
- **Version:** 0.7
- **Source summary:** Language model tools (from CMU Sphinx)
- **Homepage:** <https://cmusphinx.sourceforge.net/>
- **Last updated:** 2026-07-18T17:49:11+09:00
- **Generated:** 2026-08-03T19:37:03+00:00

## Executables

- evallm (alias)
- idngram2lm (alias)
- idngram2stats (alias)
- lm_combine (alias)
- lm_interpolate (alias)
- mergeidngram (alias)
- ngram2mgram (alias)
- text2idngram (alias)
- text2wfreq (alias)
- text2wngram (alias)
- wfreq2vocab (alias)
- wngram2idngram (alias)

## Install behavior

- Bottle: not available

## Freshness

- Page generated: 2026-08-03
- Package-manager version: 0.7
## Project history and usage

CMUCLMTK is the CMU-Cambridge statistical language modeling toolkit as carried in the CMUSphinx tree: a set of Unix-style tools for building word-frequency lists, vocabularies, n-gram counts, and backoff bigram/trigram language models for speech-recognition workflows.

### Project history

Roni Rosenfeld released version 1 of the CMU Statistical Language Modeling Toolkit in 1994, and Philip Clarkson and Rosenfeld published the CMU-Cambridge toolkit paper in 1997 describing version 2. CMUSphinx later kept the toolkit in its Subversion tree, with CMUSphinx documentation noting that no official toolkit release had appeared since the Cambridge release and that a planned CMUSphinx release would effectively be version 2.1.

### Adoption history

The 1997 paper reported use in more than 40 academic, government, and industrial laboratories in more than 12 countries. In package-manager culture it survives as a niche speech/NLP utility: this batch input records Homebrew, MacPorts, and openSUSE/zypper package names.

### How it is used

Typical use is a pipeline of small commands such as text2wfreq, wfreq2vocab, text2idngram, idngram2lm, evallm, lm_combine, and lm_interpolate to turn text corpora into n-gram language models and evaluate or combine them.

### Why package nerds care

Package nerds care because it is a fossilized but useful research-toolchain package: old speech-recognition infrastructure, many tiny command-line filters, and CMUSphinx lineage preserved in modern package managers long after the research ecosystem moved toward newer language-modeling tools.

### Timeline

- 1994: Version 1 of the CMU SLM Toolkit was released.
- 1997: Clarkson and Rosenfeld published the CMU-Cambridge toolkit paper for version 2.
- CMUSphinx era: The toolkit was maintained under the CMUSphinx Subversion tree with planned version 2.1 work.

### Related projects

- CMUSphinx and PocketSphinx are the speech-recognition projects around which this toolkit is distributed.
- SRILM is a later, widely cited language-modeling toolkit in the same n-gram research-tooling family.

### Sources

- <https://www.cs.cmu.edu/~dorcas/toolkit_documentation.html>
- <https://www.isca-archive.org/eurospeech_1997/clarkson97_eurospeech.html>
- <https://cmusphinx.github.io/wiki/cmuclmtkdevelopment>


## Security Notes

narrow executable package without higher-risk signals.

- **Geiger risk:** green / low
- narrow executable package without higher-risk signals

## Other Package-Manager Records

- zypper - cmuclmtk - 0.7-12.7: normalized package name match | openSUSE Tumbleweed package metadata: cmuclmtk from https://download.opensuse.org/tumbleweed/repo/oss/repodata/50b07339cb64c8ed4091bdbabddadc1ff5737b090e478818a195b40d8a3292861a879139b4a3987c31109699fde9fbf4a716367ddf4eef77da75f96e3193d6ed-primary.xml.zst | CMU-Cambridge Statistical Language Modeling toolkit | http://cmusphinx.sourceforge.net
- zypper - libcmuclmtk-devel - 0.7-12.7: normalized package name match | openSUSE Tumbleweed package metadata: libcmuclmtk-devel from https://download.opensuse.org/tumbleweed/repo/oss/repodata/50b07339cb64c8ed4091bdbabddadc1ff5737b090e478818a195b40d8a3292861a879139b4a3987c31109699fde9fbf4a716367ddf4eef77da75f96e3193d6ed-primary.xml.zst | CMU-Cambridge Statistical Language Modeling toolkit | http://cmusphinx.sourceforge.net
- zypper - libcmuclmtk0 - 0.7-12.7: normalized package name match | openSUSE Tumbleweed package metadata: libcmuclmtk0 from https://download.opensuse.org/tumbleweed/repo/oss/repodata/50b07339cb64c8ed4091bdbabddadc1ff5737b090e478818a195b40d8a3292861a879139b4a3987c31109699fde9fbf4a716367ddf4eef77da75f96e3193d6ed-primary.xml.zst | CMU-Cambridge Statistical Language Modeling toolkit | http://cmusphinx.sourceforge.net
- MacPorts - cmuclmtk: normalized package name match | MacPorts ports tree: audio/cmuclmtk/Portfile from https://api.github.com/repos/macports/macports-ports/git/trees/master?recursive=1


## Combined YAML source

View the package source record on GitHub. [combined/cmuclmtk.yml](https://github.com/mxcl/pkgdb/blob/main/combined/cmuclmtk.yml)


## Sources

- pkg.so package database
- Geiger risk classifier
- curated package history
- pkgdb category and tag curation
- external package-manager database matches
- cross-ecosystem install command graph