# ali-vilab/VACE

[ICCV 2025] Official implementations for paper: VACE: All-in-One Video Creation and Editing

Repository: https://github.com/ali-vilab/VACE
Canonical: https://ross.abutalabs.com/products/vace
Homepage: https://ali-vilab.github.io/VACE-Page/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: video-editing, video-generation
Last push: 2025-10-17T05:46:57+00:00

## Health v2 (maintenance only)
Score: 41/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 47, release rhythm 35, longevity 38
- inputs: {"age_days": 543, "days_push": 320, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3934, forks 280 (observed 2026-08-28T04:08:30.055200+00:00)

## What it is
VACE is the official implementation of an all-in-one video creation and editing model from Tongyi Lab, built on Wan2.1 diffusion models. It supports reference-to-video (R2V), video-to-video (V2V), and masked video-to-video (MV2V) tasks that can be freely composed.

## Use cases
- generate videos from reference images and text prompts
- edit existing videos with text instructions
- inpaint or replace objects in videos using masks
- animate a still image into a video
- extend or outpaint video frames
- swap or move objects within a video

## When to choose
- you need a single unified model for multiple video generation and editing tasks
- you want to run state-of-the-art open-source video editing on your own GPU
- you are doing research on controllable video diffusion models

## When to avoid
- you need fast real-time video editing without a high-end GPU
- you want a polished end-user GUI application rather than Python code and model weights
- your use case is simple trimming or cutting that a standard editor handles

## Facets
- artifact type: library
- maturity: active
- function: video-processing, machine-learning, deep-learning, image-processing
- domain: artificial-intelligence, deep-learning, media
- platform: python, cross-platform
- tags: video-generation, video-editing, diffusion-model, text-to-video, wan21, iccv-2025, research, video, gpu, linux

## Member repositories
- ali-vilab/VACE (main) score 41

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:30.055200+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 2026-08-29T18:24:39.427394+00:00, confidence not recorded.
  - readme: https://github.com/ali-vilab/VACE (fetched 2026-08-28T04:08:30.055200+00:00, sha 0e1cf2ae42ed)
  - homepage: https://ali-vilab.github.io/VACE-Page/ (fetched 2026-08-29T09:18:25.053683+00:00, sha d2829c7008eb)
  - registry_pypi: https://pypi.org/pypi/vace/json (fetched 2026-08-29T09:18:25.062837+00:00, sha d6cbd54c0bcf)
- Data as of 2026-08-30T08:39:29.467469+00:00.