Fork me on GitHub

Download

To install and run MARY TTS, you will need a recent Java Runtime installed (see below for version requirement).

MaryTTS 5.x

Version Release Date Java Download Size
5.2 2016-09-15 1.7+ Release notes Web installer package
Runtime package
Builder package
60 KB
49 MB
46 MB
5.2-beta3 2016-04-27 1.7+
5.2-beta2 2016-02-02 1.7+
5.2-beta1 2015-12-21 1.7+
5.1.2 2015-01-23 1.7+ Release notes Web installer package
Runtime package
Builder package
60 KB
37.8 MB
40.5 MB
5.1.1 2014-10-05 1.7+ Release notes Runtime package
Builder package
25.8 MB
30.6 MB
5.1 2014-07-16 1.7+ Release notes Runtime package
Builder package
25.8 MB
30.6 MB
5.1-beta3 2014-07-16 1.6+
5.1-beta2 2014-03-10 1.6+
5.1-beta1 2013-11-21 1.6+
5.0 2012-06-14 1.6+ Release notes Runtime package
Builder package
24.5 MB
30.2 MB

MARY TTS 4.x

Version Release Date Java Download Size
4.3.1 2011-11-30 1.6+ Release notes Standalone installer 41.2 MB
4.3.0 2010-12-20 1.5+ Release notes Standalone installer 41 MB
4.2.0 2010-12-08 1.5+ Release notes Standalone installer 41 MB
4.1.1 2010-09-09 1.5+ Release notes Standalone installer 33.9 MB
4.1.0 2010-09-01 1.5+ Release notes Standalone installer 33.8 MB
4.0.0 2009-12-18 1.5+ Release notes Standalone installer 31.4 MB
4.0-beta 2009-09-24 1.5+ Release notes

MARY 3.x

These releases are mixed open source / research license, and mostly contain native code. This makes them more difficult to reuse and port to new platforms. Support for these releases terminates with the publication of 4.0-beta.

Older comments:

DFKI has released the core MARY system as open source, including English and Tibetan synthesis and various voices, including free unit selection voices based on the CMU Arctic databases.

In addition, German synthesis is released under a DFKI research license. The system can also use MBROLA voices, under the MBROLA license.

The interactive, web-based installer which you can download below will guide you through the installation process. It will let you decide on the install location on your file system and will let you select the components you wish to install. You will need to accept the licenses appropriate for the selected components before these will actually be downloaded and installed.

The system should run on MS Windows (2000 and XP, maybe others), PC-Linux, Sun Solaris, and Mac OS X.

Version Release Date Java Download Size
3.6.0 2008-05-21 1.5+ Release notes Web installer
Standalone installer
704 KB
32.2 MB
3.5.0 2007-12-07 1.5+ Release notes
3.1.0 2007-08-17 1.4+ Release notes
3.1-beta2 2007-07-15 1.4+ Release notes
3.1-beta1 2006-12-13 1.4+ Release notes
3.0.3 2006-10-26 1.4+ Release notes
3.0.2 2006-07-04 1.4+ Release notes
3.0.1 2006-03-07 1.4+ Release notes
3.0.0 2006-02-14 1.4+ Release notes

Related downloads

Plone Speech Synthesis Tool

This Plone product endows your Plone portal with speech output capabilities. It is a client which connects to a MARY TTS server, so you will need to install the MARY TTS system to use the Plone Speech Synthesis Tool.

Installation: 1. Install MARY TTS and run the server 2. Unpack SpeechSynthesisTool in your portal’s Products/ folder; restart zope; install product using the portal_quickinstaller; and configure the location of your MARY TTS server under “site setup”->“SpeechSynthesisTool Setup”.

Download SpeechSynthesisTool-0.1.zip or get from SVN:

svn checkout https://mary.opendfki.de/repos/branches/4.x-branch/examples/client/SpeechSynthesisTool SpeechSynthesisTool