---
title: Ming-Image-0.1-Design-8bit
canonical_url: "https://www.modelscope.cn/models/mlx-community/Ming-Image-0.1-Design-8bit"
md_url: "https://www.modelscope.cn/models/mlx-community/Ming-Image-0.1-Design-8bit.md"
repository: mlx-community/Ming-Image-0.1-Design-8bit
last_updated: 2026-09-27
license: mit
pipeline_tag: text-to-image-synthesis
tasks:
  - text-to-image-synthesis
base_model:
  - inclusionAI/Ming-Image-0.1-Design
base_model_relation: finetune
parameters: 5.6B
tensor_type:
  - BF16
  - U32
  - F32
library_name:
  - mlx
  - lora
  - safetensors
  - diffusers
  - pytorch
frameworks:
  - pytorch
language:
  - en
  - zh
downloads: 35
stars: 0
tags:
  - mlx
  - safetensors
  - apple-silicon
  - text-to-image
  - diffusion
  - s3-dit
  - rgba
  - graphic-design
  - 8-bit
  - ming-image
---

# Ming-Image-0.1-Design-8bit

> Ming-Image-0.1-Design-8bit - mlx-community 在 ModelScope 开源的模型。mlx-community/Ming-Image-0.1-Design-8bit

mlx-community/Ming-Image-0.1-Design-8bit 是 ModelScope 魔搭社区上的 5.6B 参数text-to-image-synthesis模型，采用 mit 许可，基于 inclusionAI/Ming-Image-0.1-Design 构建。

- **Repository**: mlx-community/Ming-Image-0.1-Design-8bit
- **License**: mit
- **Tasks**: text-to-image-synthesis
- **Parameters**: 5.6B
- **Base model**: inclusionAI/Ming-Image-0.1-Design
- **Tags**: mlx, safetensors, apple-silicon, text-to-image, diffusion, s3-dit, rgba, graphic-design, 8-bit, ming-image
- **Downloads**: 35
- **Stars**: 0
- **Last updated**: 2026-09-27

Source: https://www.modelscope.cn/models/mlx-community/Ming-Image-0.1-Design-8bit

---

# mlx-community/Ming-Image-0.1-Design-8bit

Pre-quantized MLX tier of **[inclusionAI/Ming-Image-0.1-Design](https://huggingface.co/inclusionAI/Ming-Image-0.1-Design)** (MIT) for Apple
Silicon, loaded by the Swift/MLX port [ming-image-swift](https://github.com/xocialize/ming-image-swift). **8-bit tier.** It matches bf16 on every quality gate we run, and its memory fits a 48 GB Mac.

A Mac never has to hold the bf16 weights to use this tier. Total size: 28.3 GB; the bf16 repo is 49.8 GB. It is
part of the [Ming-Image (MLX) collection](https://huggingface.co/collections/mlx-community/ming-image-mlx-6ab8008988cbe1b217f90c61), next to
[mlx-community/Ming-Image-0.1-Design-bf16](https://huggingface.co/mlx-community/Ming-Image-0.1-Design-bf16).

## What is quantized

| Component | This tier |
|---|---|
| `mllm/`: MoE MLLM (attention, dense and shared-expert MLPs, the 256 routed experts) | 8-bit |
| `connector/`: Qwen2-1.5B connector | 8-bit |
| `transformer/`: DiT attention and feed-forward | 8-bit |
| Kept at full precision: the MoE routers, embeddings, norms, the Qwen2.5 ViT, the f32 projection heads (`mlp/`), the VAE, and the DiT's conditioning layers (adaLN, embedders, final layer) | bf16 / f32 |

Weight-only affine quantization, group size 64. The DiT never goes below 8 bits: weight-only int4 on a DiT is
real quality damage that buys no speed.

The layout is the [bf16 repo's](https://huggingface.co/mlx-community/Ming-Image-0.1-Design-bf16) upstream tree, with MLX
`.scales` / `.biases` stored beside each quantized weight and a `quantization` block in each quantized component's
`config.json`. The Swift loader reads it as published. Loading this repo gives parameters bit-identical to
quantizing the bf16 snapshot at load time (verified for every parameter).

## Quality

| Gate (same noise and scorers as bf16) | bf16 | **8-bit** |
|---|---|---|
| Render vs the fp32-truth render (1024², same noise) | 41.9 dB | **36.5 dB** |
| Text boards: exact strings · fully correct boards | 117/122 · 12/16 | **118/122 · 12/16** |
| Zones reserved for live text: clean · text-free | 19/20 · 20/20 | 19/20 · 20/20 |
| Native alpha: the recipe's usable subjects | 5/6 | 5/6 |
| Conditioning drift vs fp32 truth (learnable relL2, short / json) | 0.122 / 0.140 | 0.118 / 0.155 |

The drifted PyTorch-MPS reference renders sit 16.6 dB from the same truth render.

Full tables are in GATE-RESULTS §8 in the port's oracle (https://github.com/xocialize/ming-image-swift).

## Memory and speed

Measured on an M5 Max as process `phys_footprint`, with MLX's buffer cache capped at 2 GB (MLXEngine's default).

- Post-load resident: 7.6 GB. Peak process footprint: 27.8 GB at 1024², 2048² and 2560×1440 alike. The
  peak is the conditioning stage, when the 8-bit MLLM loads, conditions and is released.
- The 2048²-class VAE decode is bounded (chunked mid-attention plus a tiled up path, exact).
- MLXEngine declares 7.9 GB resident plus 24.4 GB activation, which is admitted on a 48 GB Mac.
- Speed at 12 steps (M5 Max): 43 s at 1024² and 244 s at 2048². bf16 takes 39 s and 245 s. Quantization saves memory,
  not time.

## Use (Swift / MLXEngine)

```swift
import MLXMingImage
import MLXToolKit

let package = MingImageT2IPackage(configuration: MingImageConfiguration(quant: .int8))
// snapshotPath nil: the engine materializes this repo; quant .int8 selects it
try await package.load()
let response = try await package.run(T2IRequest(
    prompt: "A modern tech conference poster titled 'MLX SUMMIT 2026', bold geometric shapes",
    width: 1024, height: 1024, seed: 42)) as! T2IResponse
```

Code: https://github.com/xocialize/ming-image-swift

## License

MIT, as the upstream weights. The upstream LICENSE is included.