MOSS-TTS README

The project's README from github.com/OpenMOSS/MOSS-TTS. As one markdown file: README.md.

MOSS-TTS Family


    

GitHub Trending Repository of the Day
OpenClaw

WeChat Lark

English | 简体中文

MOSS‑TTS Family is an open‑source speech and sound generation model family from MOSI.AI and the OpenMOSS team. It is designed for high‑fidelity, high‑expressiveness, and complex real‑world scenarios, covering stable long‑form speech, multi‑speaker dialogue, voice/character design, environmental sound effects, and real‑time streaming TTS.

Start here: Quickstart · Model weights · Listen to samples · Fine-tuning · Serving backends

Choose a model for your task

What you want to build Start with
Speech and voice cloning on a CPU or in a browser MOSS-TTS-Nano
Multilingual long-form narration and voice cloning MOSS-TTS-v1.5 · Local Transformer v1.5
Multi-speaker dialogue, podcasts, and dubbing MOSS-TTSD
Real-time streaming speech MOSS-TTS-Realtime
Voice design or environmental sound effects Released models · MOSS-SoundEffect v2

News

Earlier updates
  • 2026.6.2: 🚀 vLLM-Omni now supports the full MOSS-TTS series (MossTTSDelay, MossTTSRealtime, and MossTTSNano architectures), including MOSS-TTS-v1.5, MOSS-TTS, MOSS-TTSD, MOSS-SoundEffect, MOSS-VoiceGenerator, MOSS-TTS-Realtime, and MOSS-TTS-Nano. See the recipe and examples.
  • 2026.5.26: 🚀 Released MOSS-SoundEffect-v2.0, a new text-to-audio model using a DiT backbone with the Flow Matching objective, generating 48 kHz bilingual sound effects up to 30 seconds — see moss_soundeffect_v2/.
  • 2026.5.26: 🚀 Released MOSS-TTS-v1.5, with stronger multilingual synthesis when language tags are provided, more stable voice cloning, better long-reference short-text cloning, punctuation-following prosody, and explicit pause control via [pause X.Ys].
  • 2026.5.6: 🚀 MOSS-TTS and MOSS-Audio-Tokenizer now support mlx-audio. Visit the mlx-audio GitHub repository for details.
  • 2026.4.29: 📝 MOSS-TTS 2.0 is coming soon! We are collecting TTS feedback, suggestions, and feature requests via the requirements collection form.
  • 2026.4.13: 🚀 MOSS-TTS-Nano, our ~100M-parameter model, is now available! It supports multilingual voice cloning, 48 kHz stereo input/output, and streaming output on just 4 CPU cores. Check the GitHub repository and our blog for more details.
  • 2026.3.31: 📄 Our technical reports for MOSS-TTSD and MOSS-VoiceGenerator are now available on arXiv!
  • 2026.3.26: 📘 Added a tutorial on fine-tuning the MOSS-TTS-Realtime!
  • 2026.3.20: 📄 Our technical report is now available on arXiv!
  • 2026.3.18: 🚀 Added a first-class MOSS-TTS llama.cpp implementation in the companion repository OpenMOSS/llama.cpp, including end-to-end docs and a runnable pipeline for GGUF backbone inference plus ONNX audio codec decoding. See the first-class e2e guide.
  • 2026.3.16: 📘 Added a tutorial on fine-tuning the MossTTSLocal architecture, suitable for MOSS-TTS-Local-Transformer!
  • 2026.3.12: 🚀 Added SGLang backend support for the MossTTSDelay architecture, enabling efficient inference for MOSS-TTS (Delay) and MOSS-SoundEffect, with around 3× faster generation throughput!
  • 2026.3.11: 📘 Added a tutorial on fine-tuning the MossTTSDelay architecture, suitable for MOSS-TTS(Delay), MOSS-TTSD, MOSS-VoiceGenerator, and MOSS-SoundEffect!
  • 2026.3.10: ⚡️ Significantly optimized the VRAM usage of llama.cpp inference pipeline. Now 8B model fits onto 8GB GPUs!
  • 2026.3.4: 🚀 Added PyTorch-free inference support — enabling lightweight on-device deployment via llama.cpp + ONNX Runtime. Quantized GGUF weights are released at OpenMOSS-Team/MOSS-TTS-GGUF, and the ONNX audio tokenizer is available at OpenMOSS-Team/MOSS-Audio-Tokenizer-ONNX. See the llama.cpp backend for details.
  • 2026.3.4: 🎉 We add MOSS-TTS skills in ClawHub of 🦞 OpenClaw: feishu-voice-tts and moss-tts-voice.
  • 2026.2.10: 🎉🎉🎉 We have released MOSS-TTS Family. Check our Blog for more details! Our Huggingface Space is here: MOSS-TTS, MOSS-TTSD-v1.0, MOSS-VoiceGenerator.

Demo

Contents

Introduction

When a single piece of audio needs to sound like a real person, pronounce every word accurately, switch speaking styles across content, remain stable over tens of minutes, and support dialogue, role‑play, and real‑time interaction, a single TTS model is often not enough. The MOSS‑TTS Family breaks the workflow into five production‑ready models that can be used independently or composed into a complete pipeline.

Model Architecture

We train MossTTSDelay and MossTTSLocal as complementary baselines under one training/evaluation setup: Delay emphasizes long-context stability, inference speed, and production readiness, while Local emphasizes lightweight flexibility and strong objective performance for streaming-oriented systems. Together they provide reproducible references for deployment and research.

MossTTSRealtime is not a third comparison baseline; it is a capability-driven design for voice agents. By modeling multi-turn context from both prior text and user acoustics, it delivers low-latency streaming speech that stays coherent and voice-consistent across turns.

Architecture Core Mechanism Arch Details
MossTTSDelay Multi‑head parallel RVQ prediction with delay‑pattern scheduling Arch Details
MossTTSLocal Time‑synchronous RVQ blocks with a depth transformer Arch Details
MossTTSRealtime Hierarchical text–audio inputs for realtime synthesis Arch Details

Released Models

Model Architecture Size Model Card Hugging Face ModelScope
MOSS-TTS-v1.5 MossTTSDelay 8B Model Card Hugging Face ModelScope
MOSS-TTS 1.0 MossTTSDelay 8B Model Card Hugging Face ModelScope
MOSS-TTS-Local-Transformer-v1.5 MossTTSLocal 4B Model Card Hugging Face ModelScope
MOSS-TTS-Local-Transformer MossTTSLocal 1.7B Model Card Hugging Face ModelScope
MOSS‑TTSD‑V1.0 MossTTSDelay 8B Model Card Hugging Face ModelScope
MOSS‑VoiceGenerator MossTTSDelay 1.7B Model Card Hugging Face ModelScope
MOSS‑SoundEffect MossTTSDelay 8B Model Card Hugging Face ModelScope
MOSS‑SoundEffect‑v2.0 MossSoundEffectPipeline 1.3B DiT Model Card Hugging Face ModelScope
MOSS‑TTS‑Realtime MossTTSRealtime 1.7B Model Card Hugging Face ModelScope

Supported Languages

MOSS-TTS-v1.5 and MOSS-TTS-Local-Transformer-v1.5 currently support 31 languages. They keep the 20 languages supported by MOSS-TTS 1.0 and extend multilingual con

MOSS-TTS (OpenMOSS/MOSS-TTS)

Searching Database…

Что это?MOSS-TTS Family — семейство open-source моделей генерации речи и звуков от OpenMOSS, охватывающее длинное повествование, диалоги нескольких спикеров, дизайн голоса, звуковые эффекты и потоковый синтез речи в реальном времени.
Это одна модель или семейство разных моделей?Это семейство из пяти специализированных моделей, которые выбираются под задачу: MOSS-TTS для клонирования голоса/повествования, MOSS-TTSD для диалогов, MOSS-VoiceGenerator для дизайна голоса, MOSS-TTS-Realtime для голосовых агентов и MOSS-SoundEffect для звуковых эффектов.
Как это запустить?Настройте окружение Conda или uv, при желании установите FlashAttention 2, затем скачайте веса модели с Hugging Face или ModelScope и запустите quickstart-скрипты; также доступен бэкенд llama.cpp без torch с весами GGUF.
Какой бэкенд использовать для быстрого обслуживания?Для более быстрого инференса в production можно использовать бэкенды SGLang-Omni или vLLM-Omni, которые поддерживают всю серию MOSS-TTS с OpenAI-совместимым эндпоинтом /v1/audio/speech, стримингом и клонированием голоса.
Это работает полностью локально, без отправки аудио на сервер?Да, модели с открытыми весами и рассчитаны на self-hosted инференс, включая лёгкое развёртывание на устройстве через llama.cpp и ONNX Runtime.
Почему MOSS-TTS-Realtime хорош для голосовых агентов?Он использует многошаговый инкрементальный синтез с учётом контекста и Time-To-First-Byte около 180 мс, поэтому ответы в паре с текстовой LLM звучат естественно и с низкой задержкой.
Могу ли я сам дообучить эти модели?Да, в репозитории есть туториалы по дообучению обеих архитектур — MossTTSDelay и MossTTSLocal, охватывающие MOSS-TTS, MOSS-TTSD, MOSS-VoiceGenerator и MOSS-SoundEffect.
Он работает на Mac с Apple Silicon?Да, MOSS-TTS и MOSS-Audio-Tokenizer поддерживают фреймворк mlx-audio для работы на Apple Silicon.
Сколько языков он поддерживает и какое качество звука?MOSS-TTS-v1.5 и MOSS-TTS-Local-Transformer-v1.5 поддерживают 31 язык, а новейший MOSS-Audio-Tokenizer-v2 нативно выдаёт стереозвук 48 кГц.
Creator
OpenMOSS
Size
11.6 MB
Stars
4.2K
Updated
CommitSep 06
Age
8mo
Commits · 90 Days