Back to Gap Map Model components / Speech & audio

Coqui TTS

Idiap Research Institute
open source / Overall score: 2.6

A Python library for training and running text-to-speech and voice-conversion models, among them XTTS, VITS, Tacotron and a range of vocoders, with pretrained voices and support for Meta's MMS models in about 1,100 languages. Idiap Research Institute maintains it as the coqui-tts package, a fork of Coqui's original repository, which is no longer maintained.

Openness

5 high confidence
5.0
license
MPL-2.0 (OSI)
source
public
core features withheld
no — maintained by a research institute; no paid tier, enterprise edition or license key

Coqui TTS is MPL-2.0 and the public repository is the whole library, training code included. Idiap maintains it as a research institute and sells nothing beside it; some of the voices it runs, such as XTTS, carry their own non-commercial terms.

Adoption

2 high confidence
2.0

PyPI downloads of the coqui-tts package over the trailing month.

Capability

3 medium confidence
3.0

A complete synthesis toolkit with training support, level with Piper, but for one task: it does not recognize speech.

  • https://raw.githubusercontent.com/idiap/coqui-ai-TTS/HEAD/README.md recorded 2026-11-12

    README: "Pretrained models in +1100 languages."; "Tools for training new models and fine-tuning existing models in any language."; "High-performance text-to-speech and voice conversion models"; model implementations listed under spectrogram models, end-to-end models (XTTS, VITS), vocoders and voice conversion.

Verified 2026-11-12, except some scores, still awaiting re-confirmation