OpenMOSS/MOSS-TTS is an open-source Python-based family of TTS/sound generation models including multiple architectures (MossTTSDelay, MossTTSLocal, MossTTSRealtime) and released models like MOSS-TTS-v1.5. It emphasizes long-form speech, multilingual synthesis, and voice cloning, with backend options and tokenizers.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
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.
How it works
The project trains 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. MossTTSRealtime handles multi-turn context for real-time synthesis. Model architectures include MossTTSDelay (Multi‑head parallel RVQ with delay‑pattern scheduling), MossTTSLocal (Time‑synchronous RVQ blocks with a depth transformer), and MossTTSRealtime (Hierarchical text–audio inputs for realtime synthesis).
Getting started
The repository provides releases and model cards for several variants:
- MOSS-TTS-v1.5 (8B) – MossTTSDelay, with Model Card, Hugging Face, ModelScope links.
- MOSS-TTS 1.0 (8B) – MossTTSDelay, with Model Card, Hugging Face, ModelScope links.
- MOSS-TTS-Local-Transformer-v1.5 (4B) – MossTTSLocal, with Model Card, Hugging Face, ModelScope links.
- MOSS-TTS-Local-Transformer (1.7B) – MossTTSLocal, with Model Card, Hugging Face, ModelScope links.
Recent releases
RELEASES (latest 0):
- none
Traction
stars_7d: 3950 stars_1d: 3950
Behind the repo
Not provided in the facts block.
Caveats
License: Apache-2.0. Language: Python. Created: 2026-02-07. Last push: 2026-07-26. Issues: 13.