README.md

---
license: apache-2.0
library_name: transformers
pipeline_tag: image-text-to-text
base_model:
- mistral-community/pixtral-12b
base_model_relation: quantized
---

# pixtral-12b-int4-ov

 * Model creator: [mistral-community](https://huggingface.co/mistral-community)
 * Original model: [mistral-community/pixtral-12b](https://huggingface.co/mistral-community/pixtral-12b)

## Description

This is [mistral-community/pixtral-12b](https://huggingface.co/mistral-community/pixtral-12b) model converted to the [OpenVINO™ IR](https://docs.openvino.ai/2025/documentation/openvino-ir-format.html) (Intermediate Representation) format with weights compressed to INT4 by [NNCF](https://github.com/openvinotoolkit/nncf).


## Quantization Parameters

Weight compression was performed using `nncf.compress_weights` with the following parameters:

* mode: **INT4_ASYM**

## Compatibility

The provided OpenVINO™ IR model is compatible with:

* OpenVINO version 2025.2.0 and higher
* Optimum Intel 1.26.0 and higher

## Running Model Inference with [Optimum Intel](https://huggingface.co/docs/optimum/intel/index)

1. Install packages required for using [Optimum Intel](https://huggingface.co/docs/optimum/intel/index) integration with the OpenVINO backend:

```
pip install --pre -U --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/pre-release openvino_tokenizers openvino

pip install git+https://github.com/huggingface/optimum-intel.git
```

2. Run model inference

```
from PIL import Image 
import requests 
from optimum.intel.openvino import OVModelForVisualCausalLM
from transformers import AutoProcessor, TextStreamer


model_id = "OpenVINO/pixtral-12b-int4-ov"

processor = AutoProcessor.from_pretrained(model_id)
ov_model = OVModelForVisualCausalLM.from_pretrained(model_id, trust_remote_code=True)

question = "What is unusual in this picture?"
messages = [
    {"role": "user", "content": [{"type": "text", "content": question}, {"type": "image"}]},
]
text = processor.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)

url = "https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/d5fbbd1a-d484-415c-88cb-9986625b7b11"
raw_image = Image.open(requests.get(url, stream=True).raw)

inputs = processor(text=text, images=[raw_image], return_tensors="pt")
streamer = TextStreamer(processor.tokenizer, skip_prompt=True, skip_special_tokens=True)

output = ov_model.generate(**inputs, do_sample=False, max_new_tokens=100, temperature=None, top_p=None, streamer=streamer)

```


## Limitations

Check the original [model card](https://huggingface.co/mistral-community/pixtral-12b) for limitations.

## Legal information

The original model is distributed under [apache-2.0](https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/apache-2.0.md) license. More details can be found in [original model card](https://huggingface.co/mistral-community/pixtral-12b).

## Disclaimer

Intel is committed to respecting human rights and avoiding causing or contributing to adverse impacts on human rights. See [Intel’s Global Human Rights Principles](https://www.intel.com/content/dam/www/central-libraries/us/en/documents/policy-human-rights.pdf). Intel’s products and software are intended only to be used in applications that do not cause or contribute to adverse impacts on human rights.