MaryTTS
MaryTTS is a free, open-source text-to-speech platform written in Java, with support for multiple languages and tools for developing voices.
What MaryTTS is
MaryTTS is a free, open-source text-to-speech platform written in Java, with support for multiple languages and tools for developing voices. Version 5.2 supports German, British and American English, French, Italian, Luxembourgish, Russian, Swedish, Telugu, and Turkish. Its toolkits can help add language support and build unit-selection or HMM-based synthesis voices. The project distributes runtime and builder packages, along with a web installer, and requires a recent Java Runtime; version 5.2 lists Java 1.7 or later. A speech synthesis interface runs in a web browser, while its documentation includes Java API references. A Plone Speech Synthesis Tool can connect a Plone portal to a MaryTTS server. Output is available in WAVE format, but the FAQ says there is no MP3 output. The runtime is under GNU LGPL v3.0, while the builder and voices have separate license terms that may affect commercial use. The project lists version 5.2 as released on 2016-09-15.
Good
- Supports ten listed languages in version 5.2.
- Provides toolkits for language and voice development.
- Includes a browser-based speech synthesis interface.
- API and WAVE output are available.
Watch out
- Requires a recent Java Runtime.
- Does not provide MP3 output.
- Builder and voice licenses may restrict commercial use.
- Version 5.2 is listed as released in 2016.
Who it suits
MaryTTS may suit developers needing multilingual text-to-speech or tools to build synthesis voices. It is also relevant to Java and Plone users who can meet its runtime requirement.
Verdict
MaryTTS offers multilingual synthesis and voice-building tools, with a browser interface and Java API documentation. Check the Java requirement, audio format limitation, and separate builder and voice licenses before choosing it.
Plans and prices
From MaryTTS's own pricing page, checked 12 Oct 2026.
| Plan | Price | What it gives you |
|---|---|---|
| MaryTTS | Free | Open-source multilingual text-to-speech platform; runtime is LGPL; builder and voices have separate licenses |
The facts
Each one read from MaryTTS's own pages, with the date it was checked.
What it does
- API
- The documentation page links to Javadoc for the Mary system Java implementation.checked 12 Oct 2026
- API access
- Yeschecked 20 Sep 2026
- Audio formats
- The FAQ says MaryTTS does not provide MP3 output and notes compressed audio support is not a planned near-term milestone.checked 12 Oct 2026
- Audio limitation
- The FAQ says MaryTTS does not provide MP3 output and that compressed audio support is not a planned near-term milestone.checked 12 Oct 2026
- Commercial use
- The FAQ says the runtime is LGPL-licensed and allows commercial use, modification, and redistribution; builder components are GPL-licensed, and voice licenses vary.checked 12 Oct 2026
- Current listed version
- The project summary lists version 5.2, and the download page dates its release to 2016-09-15.checked 12 Oct 2026
- Deployment
- The project offers runtime and builder packages, and the download page says a recent Java Runtime is required.checked 12 Oct 2026
- Developer support
- The site directs users to the MARY users mailing list for questions and provides documentation and Java API documentation.checked 12 Oct 2026
- Distribution
- The 5.2 downloads include a web installer, runtime package, and builder package.checked 12 Oct 2026
- Export formats
- WAVEchecked 20 Sep 2026
- Integration
- A Plone Speech Synthesis Tool connects a Plone portal to a MaryTTS server.checked 12 Oct 2026
- Issue reporting
- The project page says GitHub is used to manage issues, bugs, and feature requests.checked 12 Oct 2026
- Java requirement
- The download page says a recent Java Runtime is needed and lists Java 1.7 or later for version 5.2.checked 12 Oct 2026
- Languages
- Version 5.2 supports German, British and American English, French, Italian, Luxembourgish, Russian, Swedish, Telugu, and Turkish.checked 12 Oct 2026
- License
- The project license page lists GNU Lesser General Public License v3.0.checked 12 Oct 2026
- License limits
- The FAQ says marytts-builder is GPL-licensed and voices have individual licenses that may restrict commercial use.checked 12 Oct 2026
- Maintainers
- The project was originally developed by DFKI’s Language Technology Lab and Saarland University’s Institute of Phonetics, and is now maintained by the Multimodal Speech Processing Group and DFKI.checked 12 Oct 2026
- Product
- MaryTTS is an open-source, multilingual text-to-speech synthesis platform written in Java.checked 12 Oct 2026
- Project organization
- The project summary lists DFKI GmbH as the project organization.checked 12 Oct 2026
- Release information
- The download page lists MaryTTS 5.2, released 2016-09-15.checked 12 Oct 2026
Where it runs
- Browser demo
- The homepage links to a speech synthesis interface that runs directly in a web browser.checked 12 Oct 2026
- Platforms
- web,windows,macos,linux,api,self_hostedchecked 20 Sep 2026
Money
- Free plan
- Yeschecked 20 Sep 2026
The company
- Headquarters
- Kaiserslautern, Germanychecked 28 Sep 2026
Alternatives to MaryTTS
- 1
iSpeech
Free planOpen sourceAndroidiOSWebAPI8.5Free plan - 2
Voicebox
Free planOpen sourceAndroidiOSWebAPI8.5From $12 a year - 3
Resemble AI
Free planAndroidiOSWebAPI8.3From $280 a month - 4
Speechify
Free planAndroidiOSWebAPI8.3From $10 a month - 5
Voice.ai
Free planAndroidiOSWebAPI8.3From $5 a month - 6
Fliki
Free planAndroidiOSWebAPI8.2From $28 a month