---
title: Ming-3.0-audio-tiny-8bit
canonical_url: "https://www.modelscope.cn/models/mlx-community/Ming-3.0-audio-tiny-8bit"
md_url: "https://www.modelscope.cn/models/mlx-community/Ming-3.0-audio-tiny-8bit.md"
repository: mlx-community/Ming-3.0-audio-tiny-8bit
last_updated: 2026-10-20
license: mit
model_type:
  - ming_v3_audio
architectures:
  - BailingMoeV3AudioForConditionalGeneration
base_model:
  - inclusionAI/Ming-3.0-audio-tiny
base_model_relation: quantized
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 96
stars: 0
tags:
  - 8-bit
  - mlx
  - audio
---

# Ming-3.0-audio-tiny-8bit

> Ming-3.0-audio-tiny-8bit - mlx-community 在 ModelScope 开源的模型。Ming-3.0 Audio Tiny for MLX

mlx-community/Ming-3.0-audio-tiny-8bit 是 ModelScope 魔搭社区上的机器学习模型，采用 mit 许可，基于 inclusionAI/Ming-3.0-audio-tiny 构建。

- **Repository**: mlx-community/Ming-3.0-audio-tiny-8bit
- **License**: mit
- **Base model**: inclusionAI/Ming-3.0-audio-tiny
- **Tags**: 8-bit, mlx, audio
- **Downloads**: 96
- **Stars**: 0
- **Last updated**: 2026-10-20

Source: https://www.modelscope.cn/models/mlx-community/Ming-3.0-audio-tiny-8bit

---

# Ming-3.0 Audio Tiny for MLX

Converted from [inclusionAI/Ming-3.0-audio-tiny](https://huggingface.co/inclusionAI/Ming-3.0-audio-tiny).

English and Chinese audio understanding, transcription, text chat and spoken dialogue. `task='spoken_dialogue'` loads the bundled Talker lazily and produces 44.1 kHz audio. Only the thinker language backbone and text head are quantized; audio modules, routers and Talker remain floating-point. Speculative MTP decoding and streaming are not implemented. Reference speaker extraction uses the bundled native MLX CAMPPlus encoder. Its weights retain FP32 precision. No ONNX packages are required.

```python
from mlx_audio.sts import load
model = load('mlx-community/Ming-3.0-audio-tiny-8bit', strict=True)
result = model.generate("speech.wav", task="audio_understanding",
                        text="Describe what you hear.")
print(result.text)
```

Offline inference on Apple Silicon using native MLX code. The tokenizer is loaded without executing remote code.

Source checkpoint revision: `7cbe4a75c979f6ea48f1ee31998eb02bd7012233`.

Quantized from the BF16 MLX export using **8-bit affine quantization, group size 64**. The model-specific quantization policy preserves acoustic modules and MoE routers in floating-point precision. The native CAMPPlus speaker encoder and its Apache-2.0 license are bundled; speaker weights retain FP32 precision. The full BF16 Talker, text tokenizer and audio VAE are included under `talker/`. All necessary weights and tokenizers are included. No separate model downloads or ONNX packages are required. Use an MLX Audio version with Ming Tiny support.

The example repository name above may be replaced with the local export directory or your uploaded repository ID.

Validation: strictly reloaded from a relocated directory with an empty Hugging Face cache, network access blocked and ONNX packages absent. See `validation.json` for the tested tasks and weight inventory. `SHA256SUMS` covers all package files.