MaryTTS

MaryTTS is an open-source, multilingual text-to-speech (TTS) synthesis platform. It originated from the Mary project developed at the German Research Center for Artificial Intelligence (DFKI) and is maintained by a community of researchers and developers. The system is implemented in Java and designed as a modular, extensible platform that can be deployed as a standalone server or integrated into applications.

MaryTTS provides a server that can be run locally or hosted on a network. It exposes an HTTP API for text-to-speech synthesis and can also be accessed via a web-based interface. The architecture separates text processing, linguistic analysis, and acoustic synthesis into modular components, allowing users to customize or replace parts of the pipeline and to add new voices without altering core functionality. It supports a range of input formats, including plain text and SSML, and can produce audio in common formats, depending on configuration.

The project ships with several built-in voices for multiple languages, including English, German, Spanish, and others. Additional voices are available through community contributions and can be created by training on user-provided data. The open-source nature of MaryTTS encourages experimentation, research, and prototyping in TTS and related fields.

MaryTTS is distributed under an open-source license and is available from the project’s official site and its repository, along with documentation, tutorials, and example deployments.