---
title: MeloTTS-English-MLX
canonical_url: "https://www.modelscope.cn/models/mlx-community/MeloTTS-English-MLX"
md_url: "https://www.modelscope.cn/models/mlx-community/MeloTTS-English-MLX.md"
repository: mlx-community/MeloTTS-English-MLX
last_updated: 2026-04-03
license: mit
pipeline_tag: text-to-speech
tasks:
  - text-to-speech
model_type:
  - melotts
base_model:
  - myshell-ai/MeloTTS-English
base_model_relation: finetune
parameters: 51.8M
tensor_type:
  - F32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - en
downloads: 6
stars: 0
tags:
  - mlx
  - text-to-speech
  - tts
  - vits
  - melotts
---

# MeloTTS-English-MLX

> MeloTTS-English-MLX - mlx-community 在 ModelScope 开源的模型。MLX port of myshell-ai/MeloTTS-English for Apple Silicon.

mlx-community/MeloTTS-English-MLX 是 ModelScope 魔搭社区上的 51.8M 参数text-to-speech模型，采用 mit 许可，基于 myshell-ai/MeloTTS-English 构建。

- **Repository**: mlx-community/MeloTTS-English-MLX
- **License**: mit
- **Tasks**: text-to-speech
- **Parameters**: 51.8M
- **Base model**: myshell-ai/MeloTTS-English
- **Tags**: mlx, text-to-speech, tts, vits, melotts
- **Downloads**: 6
- **Stars**: 0
- **Last updated**: 2026-04-03

Source: https://www.modelscope.cn/models/mlx-community/MeloTTS-English-MLX

---

# MeloTTS-English-MLX

MLX port of [myshell-ai/MeloTTS-English](https://huggingface.co/myshell-ai/MeloTTS-English) for Apple Silicon.

## Architecture

VITS2-based end-to-end TTS with:
- Transformer text encoder with relative position attention
- Stochastic + deterministic duration predictors
- TransformerCouplingLayer normalizing flow (4 layers)
- HiFi-GAN vocoder
- BERT prosodic features (bert-base-uncased)

**Sample rate:** 44,100 Hz  
**Speakers:** EN-US, EN-BR, EN_INDIA, EN-AU, EN-Default

## Usage

```python
from mlx_audio.tts.models.melotts import Model, ModelConfig
from mlx_audio.tts.models.melotts.text import process_text
import mlx.core as mx
import json, soundfile as sf, numpy as np
from safetensors.numpy import load_file
from huggingface_hub import snapshot_download
from pathlib import Path

# Download model
model_path = Path(snapshot_download("mlx-community/MeloTTS-English-MLX"))

# Load
with open(model_path / "config.json") as f:
    config_data = json.load(f)
model = Model(ModelConfig.from_dict(config_data))
weights = load_file(str(model_path / "model.safetensors"))
model.load_weights(list({k: mx.array(v) for k, v in weights.items()}.items()))
Model.post_load_hook(model, model_path)

# Generate
for result in model.generate("Hello, this is MeloTTS running on MLX!"):
    audio = np.array(result.audio)
    sf.write("output.wav", audio, result.sample_rate)
```

## Converted from

[myshell-ai/MeloTTS-English](https://huggingface.co/myshell-ai/MeloTTS-English) using `mlx_audio.tts.models.melotts.convert`.