---
title: Audio8-TTS-Realistic-0.6b-bf16
canonical_url: "https://www.modelscope.cn/models/mlx-community/Audio8-TTS-Realistic-0.6b-bf16"
md_url: "https://www.modelscope.cn/models/mlx-community/Audio8-TTS-Realistic-0.6b-bf16.md"
repository: mlx-community/Audio8-TTS-Realistic-0.6b-bf16
last_updated: 2026-11-02
license: apache-2.0
pipeline_tag: text-to-speech
tasks:
  - text-to-speech
model_type:
  - arktts
architectures:
  - ArkttsModel
base_model:
  - Edge0/Audio8-TTS-Realistic-Air
base_model_relation: finetune
parameters: 601.2M
tensor_type:
  - BF16
  - F32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - yue
  - zh
  - nl
  - en
  - fi
  - fr
  - de
  - id
  - it
  - ja
  - ko
  - ms
  - pl
  - pt
  - ru
  - es
  - tr
downloads: 10
stars: 4
tags:
  - mlx
  - audio
  - text-to-speech
  - tts
  - voice-cloning
  - zero-shot
  - multilingual
  - arktts
---

# Audio8-TTS-Realistic-0.6b-bf16

> Audio8-TTS-Realistic-0.6b-bf16 - mlx-community 在 ModelScope 开源的模型。Audio8-TTS-Realistic-0.6b-bf16 (MLX)

mlx-community/Audio8-TTS-Realistic-0.6b-bf16 是 ModelScope 魔搭社区上的 601.2M 参数text-to-speech模型，采用 apache-2.0 许可，基于 Edge0/Audio8-TTS-Realistic-Air 构建。

- **Repository**: mlx-community/Audio8-TTS-Realistic-0.6b-bf16
- **License**: apache-2.0
- **Tasks**: text-to-speech
- **Parameters**: 601.2M
- **Base model**: Edge0/Audio8-TTS-Realistic-Air
- **Tags**: mlx, audio, text-to-speech, tts, voice-cloning, zero-shot, multilingual, arktts
- **Downloads**: 10
- **Stars**: 4
- **Last updated**: 2026-11-02

Source: https://www.modelscope.cn/models/mlx-community/Audio8-TTS-Realistic-0.6b-bf16

---

# Audio8-TTS-Realistic-0.6b-bf16 (MLX)

MLX conversion of [Edge0/Audio8-TTS-Realistic-Air](https://huggingface.co/Edge0/Audio8-TTS-Realistic-Air)
(revision `925bbd4`) for Apple Silicon. Upstream published this model as Audio8-TTS-Realistic-0.6b and
renamed it Realistic-Air on 2026-11-02; the weights are unchanged, so this conversion is the same model
under its new name. It is a 0.6B-parameter multilingual text-to-speech
model with zero-shot voice cloning and a bundled 44.1 kHz neural codec.

Realistic is an RL-tuned successor to
[Audio8-TTS-Preview-0.6b](https://huggingface.co/mlx-community/Audio8-TTS-Preview-0.6b-bf16).
The architecture, codec, tokenizer and tensor layout are the same; only the language-model
weights differ. Anything that loads the Preview bundle loads this one.

- **Language model**: bf16 (as published upstream)
- **Codec**: fp32, converted from `codec.pth` to safetensors with weight-norm folded and conv
  weights in MLX (channels-last) layout. Byte-identical to the Preview bundle's `codec.safetensors`.
- **Architecture**: `arktts`, a DualAR transformer. A 24-layer slow AR predicts one semantic
  token per frame, and a 4-layer fast AR predicts 10 codec codebooks per frame.

Conversion parity vs the PyTorch reference (fp32, CPU, transformers 4.57.3):
- unit and block outputs within 1e-3
- reference-audio codec encoding is 100% code-exact
- greedy generation is token-exact over all 129 frames of the validation utterance
- decoded waveform max-abs 7.1e-6

## Usage (mlx-audio)

Requires [mlx-audio](https://github.com/Blaizzy/mlx-audio) **≥ 0.4.7**, the first release with the
`arktts` model family.

```bash
pip install -U "mlx-audio>=0.4.7"
```

```python
import soundfile as sf
from mlx_audio.tts.utils import load

model = load("mlx-community/Audio8-TTS-Realistic-0.6b-bf16")

# zero-shot voice cloning: the reference transcript must match the reference audio
for result in model.generate(
    text="Welcome to Audio8 TTS, running on Apple Silicon.",
    ref_audio="reference.wav",
    ref_text="Transcript of the reference clip.",
    temperature=0.7, top_p=0.9, top_k=50,
):
    sf.write("output.wav", result.audio, result.sample_rate)
```

Without `ref_audio`, the model synthesizes with a default voice.

For Swift, the [mlx-audio8-tts-swift](https://github.com/xocialize/mlx-audio8-tts-swift) package
loads this bundle as `Audio8RealisticPackage`.

## Against the Preview checkpoint

These numbers come from our own A/B of this conversion against the Preview conversion. Both
cloned two English reference voices, with paired seeds, 2 draws per line, scored by
Whisper-large-v3-turbo and CAM++.

| | Preview-0.6b | Realistic-0.6b |
|---|---|---|
| error rate, en / zh / de / es / fr / ja (macro) | 0.0122 | 0.0126 |
| speaker similarity (CAM++) | 0.7212 | 0.7211 |
| pause time per line | 0.62 s | 1.52 s |

On those six languages the two models score the same on accuracy and voice similarity. Realistic
pauses more: its takes run about 13% longer, and all of the extra time is silence.

## Supported languages

- **Upstream's original 11:** Cantonese, Chinese, Dutch, English, French, German, Italian,
  Japanese, Korean, Polish, Spanish.
- **New in Realistic, measured at 1–7% error:** Finnish, Indonesian, Malay, Portuguese, Russian,
  Turkish.
- **Experimental:** Hindi, Norwegian and Swedish work but are uneven, at 10–25% error.
- **Not usable:** Arabic, Thai and Vietnamese. Thai falls into repetition loops.

These figures come from three lines per language, two voices and two seeds. Most of the
new-language lines were not checked by native speakers.

## License

Apache-2.0, following the upstream model. The upstream fine-tunes the Apache-2.0
Audio8-TTS-Preview-0.6b, whose slow AR derives from Qwen2.5-0.5B (Apache-2.0). The bundled codec
weights are the Preview's and carry the same license.