Instructions to use microsoft/Mage-VL-Cowork with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use microsoft/Mage-VL-Cowork with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="microsoft/Mage-VL-Cowork", trust_remote_code=True)
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("microsoft/Mage-VL-Cowork", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use microsoft/Mage-VL-Cowork with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "microsoft/Mage-VL-Cowork"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "microsoft/Mage-VL-Cowork",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/microsoft/Mage-VL-Cowork

SGLang

How to use microsoft/Mage-VL-Cowork with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "microsoft/Mage-VL-Cowork" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "microsoft/Mage-VL-Cowork",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "microsoft/Mage-VL-Cowork" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "microsoft/Mage-VL-Cowork",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use microsoft/Mage-VL-Cowork with Docker Model Runner:
```
docker model run hf.co/microsoft/Mage-VL-Cowork
```

Single-file CMF build for local inference (Rust, GPU/CPU)

by infosave - opened 2 days ago

Discussion

infosave

2 days ago

We packaged this model as single CMF files that run with cortiq, a Rust inference engine with no Python: the OneVision encoder at native screenshot resolution (Qwen2-VL preprocessing, 3-axis RoPE), the merger and Qwen3-4B, on Vulkan/DX12/Metal or CPU; cortiq serve gives an OpenAI-compatible endpoint with image parts and the model's tool calls.

Files: https://huggingface.co/infosave/Mage-VL-Cowork-cmf — q8_2f (5.1 GB, greedy output identical to Transformers float32) and q4tp (3.0 GB).

On an NVIDIA A40 the q8_2f file decodes 67 tok/s (q4tp 90 tok/s) and encodes a 2206×1048 screenshot (2277 tokens) in 4.8 s; on an Apple M4 (Metal) q4tp decodes 36 tok/s.

cargo install cortiq-cli
cortiq run mage-vl-cowork-q8_2f.cmf --image screen.png --prompt "Task: open the Downloads folder." --max-tokens 1024

(cortiq 0.8.51)

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment