---
library_name: mlx
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen2.5-Coder-3B-Instruct/blob/main/LICENSE
pipeline_tag: text-generation
base_model: Qwen/Qwen2.5-Coder-3B-Instruct
tags:
- mlx
- qwen2
- 4-bit
- apple-silicon
---

# tokenfires/Qwen2.5-Coder-3B-Instruct-MLX-4bit

4-bit MLX quantization of [Qwen/Qwen2.5-Coder-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-3B-Instruct)
for Apple Silicon. Quantization: affine, 4 bits, group size 64.

## Use with LM Studio

Search for `tokenfires/Qwen2.5-Coder-3B-Instruct-MLX-4bit` in the LM Studio model downloader, or open this page and choose **Use this model → LM Studio**.

## Use with mlx-lm

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

```python
from mlx_lm import load, generate

model, tokenizer = load("tokenfires/Qwen2.5-Coder-3B-Instruct-MLX-4bit")

messages = [{"role": "user", "content": "Write a Ruby method that reverses a string."}]
prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True)
response = generate(model, tokenizer, prompt=prompt, verbose=True)
```