Developers and researchers building text-to-speech and voice cloning applications.

An open source deep learning toolkit offering pretrained multilingual text-to-speech models and training pipelines.

Self-hosted Python library, runs on CPU or GPU, installable via pip.

Developed by Coqui AI from 2021 until the company shut down in 2024, community forks continue maintenance in 2026.

Provided production-quality open TTS after the commercial Coqui AI company closed, still underpins many voice tools.

Requires Python and PyTorch, with an optional GPU for faster inference and training.

Open source under MPL-2.0, self-hosted, no credentials, original company defunct but repo remains widely used, nearly 46k stars.