Name: MiniCPM V 4 5 GGUF
Author: gaianet

Overview

MiniCPM-V 2.6 (GGUF) is a compact yet powerful vision-language model optimized for edge deployment and local inference. Unlike bloated multimodal models, this version leverages GGUF quantization, making it highly accessible for developers integrating visual understanding into resource-constrained environments via llama.cpp or similar runtimes. It excels at high-resolution image understanding, OCR, and complex visual question answering without requiring massive GPU clusters. For developers, this means a significant reduction in latency and infrastructure costs while maintaining competitive performance on visual reasoning tasks. It serves as an efficient alternative to larger proprietary APIs for applications like automated document processing, accessibility tools, and real-time visual analysis.

Highlights

GGUF quantization for efficient local CPU/GPU inference
High-resolution image understanding and precise OCR capabilities
Apache-2.0 license for flexible commercial integration
Low memory footprint ideal for edge device deployment
Strong performance-to-size ratio in visual reasoning

Usage

Install

# Install Hugging Face transformers
pip install transformers torch

SDK Usage

# Load model with transformers
from transformers import AutoModel, AutoTokenizer

model = AutoModel.from_pretrained("gaianet/MiniCPM-V-4_5-GGUF")
tokenizer = AutoTokenizer.from_pretrained("gaianet/MiniCPM-V-4_5-GGUF")

Hugging Face Download

We recommend downloading the model via the Hugging Face CLI or Hub SDK.

Guidance：Before downloading, install huggingface_hub with:

Guidance

pip install -U huggingface_hub

CLI Download

Download the full repository

huggingface-cli download gaianet/MiniCPM-V-4_5-GGUF

Download a single file to a local folder (e.g. config.json into ./dir)

huggingface-cli download gaianet/MiniCPM-V-4_5-GGUF config.json --local-dir ./dir

See the official docs for more CLI options

SDK Download

# 模型下载
from huggingface_hub import snapshot_download
model_dir = snapshot_download('gaianet/MiniCPM-V-4_5-GGUF')

Git Download

Make sure git-lfs is installed first

Git Download

git lfs install
git clone https://huggingface.co/gaianet/MiniCPM-V-4_5-GGUF

To skip LFS large-file downloads, use:

Skip LFS

GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/gaianet/MiniCPM-V-4_5-GGUF

Model files are hosted on the Hugging Face Hub — download directly via HF CLI / SDK / Git, not through this site.

PyTorch / Transformers Usage

Install Transformers

pip install -U transformers torch

Load the model and run inference

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('gaianet/MiniCPM-V-4_5-GGUF')
tokenizer = AutoTokenizer.from_pretrained('gaianet/MiniCPM-V-4_5-GGUF')

Full Documentation

来源: HuggingFace

---
base_model: openbmb/MiniCPM-V-4_5
model_creator: openbmb
model_name: MiniCPM-V-4_5
quantized_by: Second State Inc.
pipeline_tag: visual-question-answering
language:

en

zh

---

MiniCPM-V-4_5-GGUF

Original Model

openbmb/MiniCPM-V-4_5

Run with Gaianet

Prompt template:

prompt template: minicpmv

Context size:

chat_ctx_size: 128000

Run with GaiaNet:

Quick start: https://docs.gaianet.ai/node-guide/quick-start

Customize your node: https://docs.gaianet.ai/node-guide/customize

*Quantized with llama.cpp b6138*