Coqui is an open-source text-to-speech and voice cloning library that lets developers run TTS models locally with no API costs. It includes pre-trained models for many languages and tools for training custom voice models. The go-to open-source TTS solution for developers and researchers.