MOSS-TTS v1.5

OpenMOSS's general text-to-speech family: 31 languages, zero-shot voice cloning, long-form synthesis, and duration/pronunciation control (225K+ downloads, plus 60K+ on the local-transformer variant). Distinct from the spoken-dialogue MOSS-TTSD — this is the standalone TTS line. Part of OpenMOSS's 2026 omni-audio push (Audio understanding, TTS, transcription/diarization, music).