---
title: GLM-4.5-Air-mlx-8bit-gs32
canonical_url: "https://www.modelscope.cn/models/mlx-community/GLM-4.5-Air-mlx-8bit-gs32"
md_url: "https://www.modelscope.cn/models/mlx-community/GLM-4.5-Air-mlx-8bit-gs32.md"
repository: mlx-community/GLM-4.5-Air-mlx-8bit-gs32
last_updated: 2025-11-12
license: mit
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - glm4_moe
architectures:
  - Glm4MoeForCausalLM
base_model:
  - zai-org/GLM-4.5-Air
base_model_relation: quantized
parameters: 17.6B
tensor_type:
  - U32
  - BF16
  - F32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - en
inference_backends:
  - "deploy_task text"
  - "sglang 0.5.2"
downloads: 63
stars: 0
tags:
  - transformers
  - mlx
---

# GLM-4.5-Air-mlx-8bit-gs32

> GLM-4.5-Air-mlx-8bit-gs32 - mlx-community 在 ModelScope 开源的模型。mlx-community/GLM-4.5-Air-mlx-8bit-gs32

mlx-community/GLM-4.5-Air-mlx-8bit-gs32 是 ModelScope 魔搭社区上的 17.6B 参数text-generation模型，采用 mit 许可，基于 zai-org/GLM-4.5-Air 构建，可用 deploy_task text、sglang 0.5.2 部署。

- **Repository**: mlx-community/GLM-4.5-Air-mlx-8bit-gs32
- **License**: mit
- **Tasks**: text-generation
- **Parameters**: 17.6B
- **Base model**: zai-org/GLM-4.5-Air
- **Inference backends**: deploy_task text, sglang 0.5.2
- **Tags**: transformers, mlx
- **Downloads**: 63
- **Stars**: 0
- **Last updated**: 2025-11-12

Source: https://www.modelscope.cn/models/mlx-community/GLM-4.5-Air-mlx-8bit-gs32

---

# mlx-community/GLM-4.5-Air-mlx-8bit-gs32

This model [mlx-community/GLM-4.5-Air-mlx-8bit-gs32](https://huggingface.co/mlx-community/GLM-4.5-Air-mlx-8bit-gs32) was
converted to MLX format from [zai-org/GLM-4.5-Air](https://huggingface.co/zai-org/GLM-4.5-Air)
using mlx-lm version **0.28.1**.

## Recipe:
* 8-bit
* group-size 32
* 9 bits per weight (bpw)

You can find more similar MLX model quants for an Apple Mac Studio at https://huggingface.co/bibproj

---

## Use with mlx

```bash
pip install mlx-lm
```

```python
from mlx_lm import load, generate

model, tokenizer = load("mlx-community/GLM-4.5-Air-mlx-8bit-gs32")

prompt = "hello"

if tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, add_generation_prompt=True
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)
```