# Mage-VL-Cowork

> Mage-VL-Cowork A Full Training Recipe to Learn Hybrid Computer Use from Evolved Harnesses

- **Organisation:** microsoft
- **Category:** Vision-Language Models, Reasoning Models
- **Parameters:** 4.7B
- **Context length:** 262,144 tokens
- **Licence:** Apache 2.0 (Commercial use OK)
- **Released:** 2026-10-30
- **Hugging Face:** 5.8K downloads (30 days), 231 likes

## Estimated memory

| Quant | ≈ GB |
|---|---|
| Q2_K | 3.15 |
| Q3_K_M | 3.89 |
| Q4_K_M | 4.44 |
| Q5_K_M | 4.92 |
| Q6_K | 5.44 |
| Q8_0 | 6.55 |
| F16 | 10.9 |

## Runs with

- [Transformers](https://github.com/huggingface/transformers)

## About

Mage-VL-Cowork A Full Training Recipe to Learn Hybrid Computer Use from Evolved Harnesses

---

One compact 4B policy that knows when to click — and when to code.

Computer-use agents (CUAs) that confine themselves to pixels pay a heavy tax: a single screenshot consumes thousands of tokens, and re-deriving structured file edits through dozens of mouse clicks discards the fastest, most verifiable channel on the machine. Conversely, giving a frontier model naive access to both GUI and terminal tools does not automatically produce hybrid competence — without an optimized execution scaffold, even frontier teachers exhibit suboptimal channel habits that transfer directly to distilled students. Worse, existing CUA post-training pipelines rely on checkpoints already heavily specialized for GUI grounding and coding.

Mage-VL-Cowork bridges this gap. Trained end-to-end from an intermediate 4B Mage-VL checkpoint through a progressive five-stage specialize-then-unify curriculum, it combines screenshot-grounded GUI interaction (computer_use) and persistent code/terminal execution (bash) inside a single vision-language policy — with no external orchestrator. Upstream of training, we evolve the teacher harness (pruning redundant action routes and equipping domain libraries) and build Mage-CUA-Forge, a workflow-first synthesis engine that generates 293,806 tasks across 6,915 atomic skills while cutting recurring per-task generation cost by 82.6%.

## Alternatives

- [Qwen/Qwen3.5-4B](https://selfhost.directory/models/qwen/qwen3.5-4b) (4.7B, Apache 2.0, ≈4.26 GB at Q4): Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
- [Qwen/Qwen3-VL-4B-Instruct](https://selfhost.directory/models/qwen/qwen3-vl-4b-instruct) (4.4B, Apache 2.0, ≈4.26 GB at Q4): Meet Qwen3-VL — the most powerful vision-language model in the Qwen series to date.
- [Qwen/Qwen3.5-4B-Base](https://selfhost.directory/models/qwen/qwen3.5-4b-base) (4.7B, Apache 2.0, ≈4.26 GB at Q4): Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance.
- [tencent/WeVisOne-VL-4B](https://selfhost.directory/models/tencent/wevisone-vl-4b) (4.8B, Apache 2.0, ≈4.49 GB at Q4): WeVisOne-VL (WeChat Vision All-in-One) is a unified vision-language model developed by WeChat Vision, Tencent Inc. Built upon Qwen3-VL, it unifies object-level perception (open-vocabulary object detection, referring localization, object counting, object pointing, and OCR) and pixel-level perception…
- [h2oai/h2o-lightning-4b](https://selfhost.directory/models/h2oai/h2o-lightning-4b) (4.5B, Apache 2.0, ≈4.19 GB at Q4): H2O-Lightning-4B is a 4B-parameter decision model from H2O.ai, built on Qwen/Qwen3.5-4B. It answers typed decision questions about a record (a document, a ticket, a policy, a conversation) and about images that come with the record (photos, screenshots, scanned documents, charts).
- [google/medgemma-1.5-4b-it](https://selfhost.directory/models/google/medgemma-1.5-4b-it) (4.3B, health-ai-developer-foundations, ≈4.11 GB at Q4): * License: The use of MedGemma is governed by the Health AI Developer Foundations terms of use. MedGemma has not been evaluated or optimized for multi-turn applications.

More: https://selfhost.directory/models/microsoft/mage-vl-cowork/alternatives

## Runs well on

- [GeForce RTX 4060](https://selfhost.directory/models/microsoft/mage-vl-cowork/on/rtx-4060): Q8_0, est. 30 tok/s
- [Apple M4 (16 GB)](https://selfhost.directory/models/microsoft/mage-vl-cowork/on/apple-m4-16gb): Q8_0, est. 13 tok/s
- [Radeon RX 6700 XT](https://selfhost.directory/models/microsoft/mage-vl-cowork/on/rx-6700-xt): F16, est. 22 tok/s
- [GeForce RTX 3060 12GB](https://selfhost.directory/models/microsoft/mage-vl-cowork/on/rtx-3060): F16, est. 21 tok/s
- [Arc B580](https://selfhost.directory/models/microsoft/mage-vl-cowork/on/arc-b580): F16, est. 26 tok/s
- [Tesla T4](https://selfhost.directory/models/microsoft/mage-vl-cowork/on/tesla-t4): F16, est. 19 tok/s

## Similar size

- [Qwen/Qwen3.5-4B](https://selfhost.directory/models/qwen/qwen3.5-4b) (4.7B)
- [google/gemma-3-4b-it](https://selfhost.directory/models/google/gemma-3-4b-it) (4.3B)
- [Qwen/Qwen3-VL-4B-Instruct](https://selfhost.directory/models/qwen/qwen3-vl-4b-instruct) (4.4B)
- [Qwen/Qwen2.5-VL-3B-Instruct](https://selfhost.directory/models/qwen/qwen2.5-vl-3b-instruct) (3.8B)
- [google/translategemma-4b-it](https://selfhost.directory/models/google/translategemma-4b-it) (5B)
- [google/medgemma-1.5-4b-it](https://selfhost.directory/models/google/medgemma-1.5-4b-it) (4.3B)

## FAQ

**How much VRAM does Mage-VL-Cowork need?**
About 4.4 GB at Q4_K_M with an 8K context. Q8_0 needs about 6.5 GB and full precision about 11 GB. Longer contexts need more memory for the KV cache.

**Can Mage-VL-Cowork run on the RTX 4060?**
Yes. It fits at Q8_0 (about 6.2 GB), at an estimated 30 tokens/s. It fits comfortably on the Apple M4 (16 GB), Radeon RX 6700 XT, RTX 3060 12GB.

**Can I use Mage-VL-Cowork commercially?**
Yes. It is released under Apache 2.0. Permissive: use, modify and sell freely; keep the notice and licence text.

**What is the context length of Mage-VL-Cowork?**
262,144 tokens (256K). The full window needs far more memory than an 8K context.

**How do I run Mage-VL-Cowork locally?**
It runs with Transformers. Each has copy-paste examples in its docs.

## Links

- huggingface: https://huggingface.co/microsoft/Mage-VL-Cowork
- homepage: https://microsoft.github.io/Mage/cowork/#demos
- github: https://github.com/microsoft/Mage

---
Source: https://selfhost.directory/models/microsoft/mage-vl-cowork