---
title: Confucius4-TTS-mlx-int8
canonical_url: "https://www.modelscope.cn/models/mlx-community/Confucius4-TTS-mlx-int8"
md_url: "https://www.modelscope.cn/models/mlx-community/Confucius4-TTS-mlx-int8.md"
repository: mlx-community/Confucius4-TTS-mlx-int8
last_updated: 2026-06-27
license: apache-2.0
pipeline_tag: text-to-speech
tasks:
  - text-to-speech
model_type:
  - confucius4
base_model:
  - netease-youdao/Confucius4-TTS
base_model_relation: finetune
parameters: 700.1M
tensor_type:
  - F32
  - U32
library_name:
  - mlx
  - pytorch
frameworks:
  - pytorch
language:
  - zh
  - en
  - ja
  - ko
  - de
  - fr
  - es
  - id
  - it
  - th
  - pt
  - ru
  - ms
  - vi
downloads: 68
stars: 2
tags:
  - mlx
  - apple-silicon
  - text-to-speech
  - voice-cloning
  - quantized
  - int8
---

# Confucius4-TTS-mlx-int8

> Confucius4-TTS-mlx-int8 - mlx-community 在 ModelScope 开源的模型。Confucius4-TTS-mlx (int8)

mlx-community/Confucius4-TTS-mlx-int8 是 ModelScope 魔搭社区上的 700.1M 参数text-to-speech模型，采用 apache-2.0 许可，基于 netease-youdao/Confucius4-TTS 构建。

- **Repository**: mlx-community/Confucius4-TTS-mlx-int8
- **License**: apache-2.0
- **Tasks**: text-to-speech
- **Parameters**: 700.1M
- **Base model**: netease-youdao/Confucius4-TTS
- **Tags**: mlx, apple-silicon, text-to-speech, voice-cloning, quantized, int8
- **Downloads**: 68
- **Stars**: 2
- **Last updated**: 2026-06-27

Source: https://www.modelscope.cn/models/mlx-community/Confucius4-TTS-mlx-int8

---

# Confucius4-TTS-mlx (int8)

8-bit quantized MLX build of [netease-youdao/Confucius4-TTS](https://huggingface.co/netease-youdao/Confucius4-TTS)
(multilingual + cross-lingual zero-shot voice cloning, 14 languages: zh, en, ja,
ko, de, fr, es, id, it, th, pt, ru, ms, vi) for Apple Silicon.

8-bit (group 64) on the T2S body matmuls and the w2v-bert encoder linears;
`semantic_head` + norms + embeddings kept fp32 (8-bit on the head audibly
degrades pronunciation). S2A flow + BigVGAN vocoder are fp32. ~2.6 GB total.

- **T2S:** ~2.64 GB (fp32) -> ~1.2 GB
- **w2v-bert:** ~1.5 GB (fp32) -> ~0.6 GB
- **Speed (Apple M5):** RTF ~1.7 (vs ~2.4 fp32)
- **Quality:** matched to fp32 in listening tests

## Usage

Needs the `confucius4` model in [mlx-audio](https://github.com/Blaizzy/mlx-audio)
(PR [#799](https://github.com/Blaizzy/mlx-audio/pull/799)):

```python
from mlx_audio.tts.utils import load
model = load("beyoru/Confucius4-TTS-mlx-int8")
for r in model.generate("Xin chào", ref_audio="voice.wav", lang="vi"):
    ...  # r.audio at 22050 Hz
```

## Attribution & license
- Model & architecture: netease-youdao/Confucius4-TTS (Apache-2.0)
- Vocoder: NVIDIA BigVGAN v2; speaker encoder: 3D-Speaker CAMPPlus (funasr)
- MLX port by [Hert4](https://github.com/Hert4), released under Apache-2.0.