---
title: Qwen-Image-2512-6bit-MLX
canonical_url: "https://www.modelscope.cn/models/themindstudio/Qwen-Image-2512-6bit-MLX"
md_url: "https://www.modelscope.cn/models/themindstudio/Qwen-Image-2512-6bit-MLX.md"
repository: themindstudio/Qwen-Image-2512-6bit-MLX
last_updated: 2026-05-01
license: apache-2.0
base_model:
  - Qwen/Qwen-Image-2512
base_model_relation: finetune
parameters: 11.7B
tensor_type:
  - F32
  - U32
  - BF16
library_name:
  - lora
  - mlx
  - safetensors
downloads: 411
stars: 0
tags:
  - text-to-image
  - mlx
  - qwen
  - 6-bit
  - flux
  - image-generation
---

# Qwen-Image-2512-6bit-MLX

> Qwen-Image-2512-6bit-MLX - themindstudio 在 ModelScope 开源的模型。6-bit quantized version of the Qwen-Image-2512 model

themindstudio/Qwen-Image-2512-6bit-MLX 是 ModelScope 魔搭社区上的 11.7B 参数机器学习模型，采用 apache-2.0 许可，基于 Qwen/Qwen-Image-2512 构建。

- **Repository**: themindstudio/Qwen-Image-2512-6bit-MLX
- **License**: apache-2.0
- **Parameters**: 11.7B
- **Base model**: Qwen/Qwen-Image-2512
- **Tags**: text-to-image, mlx, qwen, 6-bit, flux, image-generation
- **Downloads**: 411
- **Stars**: 0
- **Last updated**: 2026-05-01

Source: https://www.modelscope.cn/models/themindstudio/Qwen-Image-2512-6bit-MLX

---

# Qwen-Image-2512-6bit-MLX

This repository contains a **6-bit quantized** version of the [**Qwen-Image-2512**](https://huggingface.co/Qwen/Qwen-Image-2512) model, converted to the **MLX** format for high-performance inference on Apple Silicon

## Model Description

**Qwen-Image-2512** is a state-of-the-art text-to-image foundational model developed by the Qwen team. It features significant improvements in:
- **Human Realism**: Enhanced skin textures and anatomical accuracy.
- **Natural Detail**: Superior rendering of landscapes, animals, and intricate textures.
- **Text Rendering**: High accuracy in generating readable text within images.

This MLX version has been quantized to **6-bit** to provide a balance between generation quality and memory efficiency, making it ideal for Macs with 32GB of RAM or more.

## Usage 

To use this model, download [MindCraft Studio](https://themindstudio.cc/mindcraft) and import this as a base model.