# microsoft/Mage

Repository: https://github.com/microsoft/Mage
Canonical: https://ross.abutalabs.com/products/microsoft-mage
Language: Python
License: MIT
License Family: permissive
Last push: 2026-08-10T09:21:40+00:00

## Health v2 (maintenance only)
Score: 57/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 97, release rhythm 35, longevity 5
- inputs: {"age_days": 79, "days_push": 23, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1516, forks 164 (observed 2026-08-28T04:04:56.428660+00:00)

## What it is
Mage is a family of lightweight 4B-parameter multimodal models from Microsoft, including Mage-VL for image and video understanding and Mage-Flow for text-to-image generation and instruction-based editing. It is designed for research-friendly training, fine-tuning, and deployment on modest hardware.

## Use cases
- run a small vision-language model for image and video understanding
- generate images from text prompts with a compact model
- edit images via natural language instructions
- fine-tune a multimodal model on a limited compute budget
- experiment with post-training of vision-language models
- build vertical-domain applications with multimodal understanding

## When to choose
- you need a compact multimodal model that trains or fine-tunes on modest hardware
- you want both understanding and generation models from one consistent family
- you are doing controlled research experiments at a fixed parameter budget

## When to avoid
- you need state-of-the-art quality regardless of model size
- you need a production-ready turnkey multimodal API rather than a research codebase
- you work outside Python/GPU environments

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, llm-inference, llm-training, image-processing, video-processing, stable-diffusion
- domain: artificial-intelligence, machine-learning, deep-learning, computer-vision, image-processing, large-language-models
- platform: python, cross-platform
- tags: multimodal, vision-language-model, text-to-image, image-editing, 4b-parameters, research-models, fine-tuning, microsoft, video, natural-language-processing, gpu, linux

## Member repositories
- microsoft/Mage (main) score 57

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:56.428660+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:32:06.787946+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/Mage (fetched 2026-08-28T04:04:56.428660+00:00, sha 3d655261f142)
- Data as of 2026-08-30T08:39:29.467469+00:00.