---
title: MOSS-TTSD-v1.0-MLX-8bit
canonical_url: "https://www.modelscope.cn/models/mlx-community/MOSS-TTSD-v1.0-MLX-8bit"
md_url: "https://www.modelscope.cn/models/mlx-community/MOSS-TTSD-v1.0-MLX-8bit.md"
repository: mlx-community/MOSS-TTSD-v1.0-MLX-8bit
last_updated: 2026-04-02
license: apache-2.0
pipeline_tag: text-to-speech
tasks:
  - text-to-speech
base_model:
  - OpenMOSS-Team/MOSS-TTSD-v1.0
base_model_relation: quantized
parameters: 2.4B
tensor_type:
  - BF16
  - U32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - zh
  - en
downloads: 38
stars: 1
tags:
  - mlx
  - tts
  - speech
  - multi-speaker
  - dialogue
  - apple-silicon
  - quantized
  - 8bit
---

# MOSS-TTSD-v1.0-MLX-8bit

> MOSS-TTSD-v1.0-MLX-8bit - mlx-community 在 ModelScope 开源的模型。MOSS TTSD v1.0 — MLX 8-bit

mlx-community/MOSS-TTSD-v1.0-MLX-8bit 是 ModelScope 魔搭社区上的 2.4B 参数text-to-speech模型，采用 apache-2.0 许可，基于 OpenMOSS-Team/MOSS-TTSD-v1.0 构建。

- **Repository**: mlx-community/MOSS-TTSD-v1.0-MLX-8bit
- **License**: apache-2.0
- **Tasks**: text-to-speech
- **Parameters**: 2.4B
- **Base model**: OpenMOSS-Team/MOSS-TTSD-v1.0
- **Tags**: mlx, tts, speech, multi-speaker, dialogue, apple-silicon, quantized, 8bit
- **Downloads**: 38
- **Stars**: 1
- **Last updated**: 2026-04-02

Source: https://www.modelscope.cn/models/mlx-community/MOSS-TTSD-v1.0-MLX-8bit

---

# MOSS TTSD v1.0 — MLX 8-bit

This repository contains an MLX-native int8 conversion of MOSS TTSD v1.0 for multi-speaker dialogue generation on Apple Silicon.

> Note
> This repo is a community mirror of the canonical MLX conversion maintained by
> [AppAutomaton](https://github.com/appautomaton) at
> [`appautomaton/openmoss-ttsd-mlx`](https://huggingface.co/appautomaton/openmoss-ttsd-mlx).

## Variants

| Path | Precision |
| --- | --- |
| `mlx-int8/` | int8 quantized weights |

## Model Details

- Developed by: AppAutomaton
- Shared by: `mlx-community`
- Original MLX repo: [`appautomaton/openmoss-ttsd-mlx`](https://huggingface.co/appautomaton/openmoss-ttsd-mlx)
- Upstream model: [`OpenMOSS-Team/MOSS-TTSD-v1.0`](https://huggingface.co/OpenMOSS-Team/MOSS-TTSD-v1.0)
- Task: multi-speaker text-to-speech
- Runtime: MLX on Apple Silicon

## How to Get Started

Command-line generation with [`mlx-speech`](https://github.com/appautomaton/mlx-speech):

```bash
python scripts/generate_moss_ttsd.py \
  --text "[S1] Watson, I think we should go. [S2] Give me one moment." \
  --output outputs/dialogue.wav
```

Minimal Python usage:

```python
from mlx_speech.generation import MossTTSDelayModel

model = MossTTSDelayModel.from_path("mlx-int8")
```

Speaker turns are tagged with `[S1]` and `[S2]` in the input text.

## Notes

- This repo contains the quantized MLX runtime artifact only.
- The conversion keeps the original TTSD architecture and remaps weights explicitly for MLX inference.
- The current runtime path is designed around speaker-tagged dialogue input and shared codec decoding.
- This mirror is a duplicated repo, not an automatically synchronized namespace mirror.

## Links

- Canonical MLX repo: [`appautomaton/openmoss-ttsd-mlx`](https://huggingface.co/appautomaton/openmoss-ttsd-mlx)
- Source code: [`mlx-speech`](https://github.com/appautomaton/mlx-speech)
- More examples: [AppAutomaton](https://github.com/appautomaton)

## License

Apache 2.0 — following the upstream license published with [`OpenMOSS-Team/MOSS-TTSD-v1.0`](https://huggingface.co/OpenMOSS-Team/MOSS-TTSD-v1.0).