---
title: Lens-Turbo-mlx-6bit
canonical_url: "https://www.modelscope.cn/models/themindstudio/Lens-Turbo-mlx-6bit"
md_url: "https://www.modelscope.cn/models/themindstudio/Lens-Turbo-mlx-6bit.md"
repository: themindstudio/Lens-Turbo-mlx-6bit
last_updated: 2026-06-19
license: mit
pipeline_tag: text-to-image-synthesis
tasks:
  - text-to-image-synthesis
base_model:
  - microsoft/Lens-Turbo
base_model_relation: finetune
parameters: 12.9B
tensor_type:
  - U8
  - F32
  - BF16
  - U32
library_name:
  - mlx
  - lora
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 43
stars: 0
tags:
  - mlx
  - text-to-image
  - lens-turbo
  - image-generation
---

# Lens-Turbo-mlx-6bit

> Lens-Turbo-mlx-6bit - themindstudio 在 ModelScope 开源的模型。This repository contains the 6-bit quantized version of Microsoft's Lens-Turbo text-to-image model, optimized for Apple Silicon using MLX.

themindstudio/Lens-Turbo-mlx-6bit 是 ModelScope 魔搭社区上的 12.9B 参数text-to-image-synthesis模型，采用 mit 许可，基于 microsoft/Lens-Turbo 构建。

- **Repository**: themindstudio/Lens-Turbo-mlx-6bit
- **License**: mit
- **Tasks**: text-to-image-synthesis
- **Parameters**: 12.9B
- **Base model**: microsoft/Lens-Turbo
- **Tags**: mlx, text-to-image, lens-turbo, image-generation
- **Downloads**: 43
- **Stars**: 0
- **Last updated**: 2026-06-19

Source: https://www.modelscope.cn/models/themindstudio/Lens-Turbo-mlx-6bit

---

-
# Lens-Turbo-MLX-6bit

This repository contains the 6-bit quantized version of **Microsoft's Lens-Turbo** text-to-image model, optimized for Apple Silicon using MLX.

Lens-Turbo is a fast, high-quality 3.8B parameter text-to-image model. 

## Model Details
- **Original Model**: [microsoft/Lens-Turbo](https://huggingface.co/microsoft/Lens-Turbo)
- **MindCraft Studio**: [MindCraft Studio Homepage](https://themindstudio.cc/mindcraft)
- **Quantization**: 6-bit weight-only quantization (`quantize=6`) via MLX.
- **Inference Steps**: Recommended 4 steps.
- **Guidance Scale**: 1.0 (CFG is disabled/forced to 1.0 for the distilled Turbo architecture).

## How to Use in MindCraft Studio
1. Open [MindCraft Studio](https://themindstudio.cc/mindcraft).
2. Go to **Settings** -> **Model Library** -> **Import Local Model**.
3. Point to this directory.
4. The model will show up as `Lens` (Turbo) and is ready for high-speed generation.

## Credits & References
- **Original Model**: [microsoft/Lens-Turbo](https://huggingface.co/microsoft/Lens-Turbo)
- **MindCraft Studio**: [themindstudio.cc/mindcraft](https://themindstudio.cc/mindcraft)