Mini toolsMOSS TTS v1.5
By OpenMOSS
MOSS-TTS is an open-source audio-generation model family from MOSI.AI and the OpenMOSS team. It is designed for high-fidelity, expressive, production-style speech and sound workflows, including stable long-form speech, multilingual and code-switched synthesis, multi-speaker dialogue, voice or character design, low-latency voice agents, and sound-effect generation. The family includes MOSS-TTS for zero-shot voice cloning and long speech, MOSS-TTSD for expressive multi-speaker dialogue, MOSS-VoiceGenerator for text-prompted voice design without reference speech, MOSS-TTS-Realtime for low-latency multi-turn voice agents, and MOSS-SoundEffect for controllable environmental and action sounds. It also supports lightweight deployment paths such as llama.cpp plus ONNX Runtime, GGUF weights, and low-memory configurations for smaller GPUs.
New
Multimodal
Overview
MOSS-TTS is OpenMOSS and MOSI.AI’s open-source speech and sound generation family for voice cloning, dialogue speech, voice design, real-time TTS, and sound effects.
Last updated: July 7, 2026
