Mozilla TTS is an open-source text-to-speech library from Mozilla that turns written text into natural-sounding speech using deep learning. Published on GitHub under the Mozilla Public License 2.0, it has earned more than 10,100 stars and was last actively updated in 2023. The project grew out of Mozilla’s voice research work and became one of the most widely used open TTS codebases for both research and production voice products.
TTS is built on the latest speech synthesis research and was designed to strike the best trade-off among ease of training, speed, and output quality. It ships with pretrained models and has been used in more than 20 languages across products and research projects. The library covers the full pipeline: acoustic models such as Tacotron and Tacotron 2, vocoders including MelGAN, Multi-Band MelGAN, and Glow-TTS flows, plus a speaker encoder for voice cloning style workflows. Tools for measuring dataset quality are included, so teams can clean and validate their own recordings before training. Everything runs in Python with PyTorch and TensorFlow 2 support.
Mozilla TTS is completely free and open source under the MPL-2.0 license. You can download it from GitHub or PyPI, train models on your own hardware, and use the results in products and research without license fees. The only costs are your own compute and data preparation time.
Mozilla TTS suits developers and researchers who need a proven, hackable speech synthesis stack: teams building voice assistants, accessibility tools, audiobook pipelines, or multilingual narration features, and academics experimenting with new TTS architectures. If you want full control over training data and voices instead of a hosted API, this library is a strong starting point.