Microsoft
Mage VL Cowork
Microsoft's 4.7B-parameter open vision-language model with a 256K-token context, ready to serve on Runpod Serverless with vLLM or on a Pod.
Mage VL Cowork
Specifications
Provider
Microsoft
Modality
Vision + language
Task
Text generation
Parameters
4.7B
Active parameters
Context
256K tokens
Precision
BF16
Weights
10
License
apache-2.0
Released
2026-10-29
Endpoint ID
Weights on Hugging Face
microsoft/Mage-VL-Cowork
Frequently asked questions
How do I deploy Mage VL Cowork on Runpod?
Deploy the vLLM worker from the Runpod Hub as a Serverless endpoint and set MODEL_NAME to microsoft/Mage-VL-Cowork. The endpoint is OpenAI-compatible and scales to zero when idle. You can also run vLLM on a Pod.
What GPU do I need to run Mage VL Cowork?
Start with RTX 4090 (24 GB VRAM). The weights take about 10 GB; longer contexts and more concurrent requests need extra memory for the KV cache.
How much does Mage VL Cowork cost on Runpod?
Serverless endpoints and Pods bill per second for the GPU you choose. A Serverless endpoint costs nothing while scaled to zero.
microsoft-mage-vl-cowork
microsoft/Mage-VL-Cowork
Vision + language
2026-10-29 17:00
RTX 4090