---
title: Macaw-OptiQ-4bit
canonical_url: "https://www.modelscope.cn/models/mlx-community/Macaw-OptiQ-4bit"
md_url: "https://www.modelscope.cn/models/mlx-community/Macaw-OptiQ-4bit.md"
repository: mlx-community/Macaw-OptiQ-4bit
last_updated: 2026-09-15
license: other
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - lfm2
architectures:
  - Lfm2ForCausalLM
base_model:
  - badtheorylabs/Macaw
base_model_relation: quantized
parameters: 540.3M
tensor_type:
  - U32
  - BF16
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - en
downloads: 13
stars: 0
tags:
  - mlx
  - optiq
  - quantized
  - 4bit
  - mixed-precision
  - lfm2
  - agent
  - tool-calling
  - on-device
  - apple-silicon
---

# Macaw-OptiQ-4bit

> Macaw-OptiQ-4bit - mlx-community 在 ModelScope 开源的模型。mlx-community/Macaw-OptiQ-4bit

mlx-community/Macaw-OptiQ-4bit 是 ModelScope 魔搭社区上的 540.3M 参数text-generation模型，采用 other 许可，基于 badtheorylabs/Macaw 构建。

- **Repository**: mlx-community/Macaw-OptiQ-4bit
- **License**: other
- **Tasks**: text-generation
- **Parameters**: 540.3M
- **Base model**: badtheorylabs/Macaw
- **Tags**: mlx, optiq, quantized, 4bit, mixed-precision, lfm2, agent, tool-calling, on-device, apple-silicon
- **Downloads**: 13
- **Stars**: 0
- **Last updated**: 2026-09-15

Source: https://www.modelscope.cn/models/mlx-community/Macaw-OptiQ-4bit

---

# mlx-community/Macaw-OptiQ-4bit

> **Built with [mlx-optiq](https://mlx-optiq.com)**, the MLX-native toolkit to quantize, fine-tune, and serve LLMs locally on Apple Silicon, no PyTorch and no cloud. [All OptiQ quants](https://mlx-optiq.com/models) · [Docs](https://mlx-optiq.com/docs/) · [LFM2.5 family](https://mlx-optiq.com/docs/lfm2.5)

An OptiQ mixed-precision quant of [badtheorylabs/Macaw](https://huggingface.co/badtheorylabs/Macaw), an on-device Mac assistant built on LFM2.5-2.6B. 1.93 GB on disk, down from 5.2 GB at bf16.

Macaw is a tool-calling agent, so the quant is aimed at keeping tool calls well formed rather than at raw benchmark scores.

## What it is

| Property | Value |
|---|---|
| Base | [badtheorylabs/Macaw](https://huggingface.co/badtheorylabs/Macaw) (LFM2.5-2.6B derivative) |
| Architecture | `lfm2` — hybrid conv + full attention, 30 layers |
| Method | OptiQ mixed-precision, per-layer bit allocation reused from the base family |
| On disk | 1.93 GB (bf16: 5.2 GB) |
| Context | 128k |

Macaw keeps its base architecture, so which layers tolerate fewer bits is unchanged. The per-layer allocation comes from [LFM2.5-2.6B-OptiQ-4bit](https://huggingface.co/mlx-community/LFM2.5-2.6B-OptiQ-4bit) rather than a fresh sensitivity sweep: 167 of 167 layers matched, 80 kept at 8-bit and 87 at 4-bit.

## Run it

```bash
pip install mlx-optiq
optiq serve --model mlx-community/Macaw-OptiQ-4bit
```

That gives you an OpenAI and Anthropic compatible endpoint with mixed-precision KV cache, tool-call healing and prompt caching. The base model's recommended sampling ships in `generation_config.json` and `optiq serve` applies it without any flags.

## Links

- **Project website:** [mlx-optiq.com](https://mlx-optiq.com/)
- **All OptiQ quants:** [mlx-optiq.com/models](https://mlx-optiq.com/models)
- **PyPI:** [pypi.org/project/mlx-optiq](https://pypi.org/project/mlx-optiq/)
- **Base model:** [badtheorylabs/Macaw](https://huggingface.co/badtheorylabs/Macaw)