richardyoung/ minicpm-v-4.6:Q4_K_M

Updated 17 hours ago

MiniCPM-V 4.6 vision model (1.3B): image understanding and OCR in about 1.6 GB. Image input works.

vision

ollama run richardyoung/minicpm-v-4.6:Q4_K_M

curl http://localhost:11434/api/chat \
  -d '{
    "model": "richardyoung/minicpm-v-4.6:Q4_K_M",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='richardyoung/minicpm-v-4.6:Q4_K_M',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'richardyoung/minicpm-v-4.6:Q4_K_M',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 17 hours ago

17 hours ago

c5d827b4c7e0 · 1.6GB

model

archqwen35

·

parameters752M

·

quantizationQ4_K_M

529MB

projector

archclip

·

parameters548M

·

quantizationF16

1.1GB

Readme

MiniCPM-V-4.6

MiniCPM-V 4.6 vision model (1.3B): image understanding and OCR in about 1.6 GB. Image input works.

🚀 Overview

GGUF build of openbmb/MiniCPM-V-4.6, OpenBMB’s compact vision-language model: describes images, answers questions about them and reads text in photos and documents. It ships with its vision projector, so image input works in Ollama.

🎯 Key Features

Image understanding and visual Q&A
OCR of photos, screenshots and documents
Tiny: about 1.6 GB with the vision projector
Apache-2.0 license

🏷️ Available Versions

Tag	Size (+ 1.11 GB projector)	BPW	Notes
latest / Q4_K_M	0.53 GB	4.85	Recommended
Q8_0	0.81 GB	8.5	Near-lossless

💻 Quick Start

ollama run richardyoung/minicpm-v-4.6
# then drag an image file into the prompt, or send it in the API "images" field

🛠️ Use Cases

Local image description and visual Q&A
Screenshot and document reading
Low-VRAM and CPU vision

✅ Verified

Tested through Ollama before upload with a synthetic invoice and a shapes image: it transcribed the whole invoice exactly and named both shapes and their colors.

🔧 Technical Details

Base Model: openbmb/MiniCPM-V-4.6
Parameters: 1.3B
Context Length: 256K tokens
Quantization: GGUF via llama.cpp; vision projector kept at F16; converted with --no-mtp (the release declares a multi-token-prediction layer it does not ship)
License: Apache-2.0 (inherited from the base model)

🙏 Acknowledgments

Base Model: OpenBMB
Quantization: llama.cpp

Built & maintained by Richard Young · DeepNeuro

# MiniCPM-V-4.6

MiniCPM-V 4.6 vision model (1.3B): image understanding and OCR in about 1.6 GB. Image input works.

## 🚀 Overview

GGUF build of [openbmb/MiniCPM-V-4.6](https://huggingface.co/openbmb/MiniCPM-V-4.6), OpenBMB's compact vision-language model: describes images, answers questions about them and reads text in photos and documents. It ships with its vision projector, so image input works in Ollama.

## 🎯 Key Features

- Image understanding and visual Q&A
- OCR of photos, screenshots and documents
- Tiny: about 1.6 GB with the vision projector
- Apache-2.0 license

## 🏷️ Available Versions

| Tag | Size (+ 1.11 GB projector) | BPW | Notes |
|---|---|---|---|
| latest / Q4_K_M | 0.53 GB | 4.85 | Recommended |
| Q8_0 | 0.81 GB | 8.5 | Near-lossless |

## 💻 Quick Start

```bash
ollama run richardyoung/minicpm-v-4.6
# then drag an image file into the prompt, or send it in the API "images" field
```

## 🛠️ Use Cases

- Local image description and visual Q&A
- Screenshot and document reading
- Low-VRAM and CPU vision

## ✅ Verified

Tested through Ollama before upload with a synthetic invoice and a shapes image: it transcribed the whole invoice exactly and named both shapes and their colors.

## 🔧 Technical Details

- **Base Model:** [openbmb/MiniCPM-V-4.6](https://huggingface.co/openbmb/MiniCPM-V-4.6)
- **Parameters:** 1.3B
- **Context Length:** 256K tokens
- **Quantization:** GGUF via llama.cpp; vision projector kept at F16; converted with `--no-mtp` (the release declares a multi-token-prediction layer it does not ship)
- **License:** Apache-2.0 (inherited from the base model)

## 🙏 Acknowledgments

- **Base Model:** OpenBMB
- **Quantization:** llama.cpp

*Built & maintained by [Richard Young](https://deepneuro.ai/richard) · DeepNeuro*

Paste, drop or click to upload images (.png, .jpeg, .jpg, .svg, .gif)