A family of fine-tuned Orpheus models by Hypa AI, designed for expressive, multi-speaker, and low-resource multilingual text-to-speech synthesis.