---
title: MiniCPM-V-4.7-1B-4bit
canonical_url: "https://www.modelscope.cn/models/mlx-community/MiniCPM-V-4.7-1B-4bit"
md_url: "https://www.modelscope.cn/models/mlx-community/MiniCPM-V-4.7-1B-4bit.md"
repository: mlx-community/MiniCPM-V-4.7-1B-4bit
last_updated: 2026-10-13
license: apache-2.0
pipeline_tag: image-text-to-text
tasks:
  - image-text-to-text
model_type:
  - minicpmv4_7
architectures:
  - MiniCPMV4_7ForConditionalGeneration
base_model:
  - openbmb/MiniCPM-V-4.7-1B
base_model_relation: quantized
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 100
stars: 0
tags:
  - mlx
  - mlx-vlm
  - minicpmv4_7
  - vision
  - multimodal
  - 4bit
---

# MiniCPM-V-4.7-1B-4bit

> MiniCPM-V-4.7-1B-4bit - mlx-community 在 ModelScope 开源的模型。MiniCPM-V-4.7-1B-4bit

mlx-community/MiniCPM-V-4.7-1B-4bit 是 ModelScope 魔搭社区上的image-text-to-text模型，采用 apache-2.0 许可，基于 openbmb/MiniCPM-V-4.7-1B 构建。

- **Repository**: mlx-community/MiniCPM-V-4.7-1B-4bit
- **License**: apache-2.0
- **Tasks**: image-text-to-text
- **Base model**: openbmb/MiniCPM-V-4.7-1B
- **Tags**: mlx, mlx-vlm, minicpmv4_7, vision, multimodal, 4bit
- **Downloads**: 100
- **Stars**: 0
- **Last updated**: 2026-10-13

Source: https://www.modelscope.cn/models/mlx-community/MiniCPM-V-4.7-1B-4bit

---

# MiniCPM-V-4.7-1B-4bit

Converted from [openbmb/MiniCPM-V-4.7-1B](https://huggingface.co/openbmb/MiniCPM-V-4.7-1B) at revision `984a89aafcd5160164f4f0d04b4be0f4eec03cb8` with [MLX-VLM main](https://github.com/Blaizzy/mlx-vlm/commit/e928101a144a3f11a27c7ef8d71b6d889b2d250c) (`e928101a144a3f11a27c7ef8d71b6d889b2d250c`, version 0.7.8). Support was added in [PR #2499](https://github.com/Blaizzy/mlx-vlm/pull/2499).

4BIT weight quantization (affine, group size 64). This is mixed precision: vision modules, visual mergers, embeddings, output heads, and linear attention remain BF16; recurrent `A_log` tensors retain their source precision. The 35B-A3B expert routers use 8-bit affine quantization.

See the [original model card](https://huggingface.co/openbmb/MiniCPM-V-4.7-1B) for the architecture, intended uses, and limitations. This conversion retains the source model's apache-2.0 license.

## Usage

Install MLX-VLM from `main` with MiniCPM-V 4.7 support:

```bash
pip install -U git+https://github.com/Blaizzy/mlx-vlm.git@main
mlx_vlm.generate --model mlx-community/MiniCPM-V-4.7-1B-4bit --image path/to/image.jpg --prompt "Describe this image." --max-tokens 200 --temperature 0
```

The native NumPy/Pillow processor requires neither PyTorch nor remote code. It supports text, images, multiple images, and video. The default visual downsampling is `16x`; pass `--processor-kwargs '{"downsample_mode": "4x"}'` for finer detail.

## Reproduce conversion

```bash
mlx_vlm.convert --hf-path openbmb/MiniCPM-V-4.7-1B --revision 984a89aafcd5160164f4f0d04b4be0f4eec03cb8 --mlx-path MiniCPM-V-4.7-1B-4bit --dtype bfloat16 -q --q-bits 4 --q-group-size 64 --q-mode affine
```

For the exact converter version, install MLX-VLM at commit `e928101a144a3f11a27c7ef8d71b6d889b2d250c`. Conversion options, per-module quantization overrides, and weight SHA-256 hashes are recorded in [conversion.json](conversion.json).

## Validation

The saved checkpoint passed strict reloading, quantization configuration checks, greedy text generation, and cat-image generation in both `16x` and `4x` downsampling modes on an Apple M3 Ultra with MLX 0.32.3. See [validation.json](validation.json) for the outputs and environment. These are smoke checks, not a quality benchmark. Floating-point 4-bit formats can differ in output quality from affine quantization.