# jd-opensource/JoyAI-Video-Edit

[Official Repo] JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion

Repository: https://github.com/jd-opensource/JoyAI-Video-Edit
Canonical: https://ross.abutalabs.com/products/joyai-video-edit
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-08-25T08:34:08+00:00

## Health v2 (maintenance only)
Score: 57/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 35, longevity 2
- inputs: {"age_days": 29, "days_push": 8, "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 1671, forks 74 (observed 2026-08-28T04:05:19.634548+00:00)

## What it is
JoyAI-Video-Edit is a real-time, instruction-guided video editing system that applies natural-language edits to live or uploaded video streams causally as frames arrive. It combines an MLLM condition encoder, causal video VAE, and a 16B-parameter multimodal diffusion transformer, reaching up to 30 FPS at 720p on high-end GPUs.

## Use cases
- edit live camera streams with natural-language instructions in real time
- apply instruction-guided edits to uploaded videos without batch processing
- run reference-image-guided video-to-video editing preserving subject identity
- stream video editing at 720p on a single consumer GPU
- deploy low-latency streaming video-to-video generation services

## When to choose
- you need real-time or streaming video editing rather than offline batch processing
- you want open-ended edits driven by natural-language instructions
- you have a modern NVIDIA GPU and need high-throughput 720p video-to-video editing

## When to avoid
- you need precise frame-accurate non-generative video editing (use a traditional NLE)
- you lack a capable GPU or cannot host a 16B-parameter model
- you need lightweight CPU-only video processing

## Facets
- artifact type: library
- maturity: active
- function: video-processing, machine-learning, deep-learning, llm-inference, image-processing
- domain: artificial-intelligence, deep-learning, media
- platform: python
- tags: video-editing, diffusion-transformer, autoregressive, real-time, instruction-guided, streaming-video, video-to-video, ml, video, gpu, linux

## Member repositories
- jd-opensource/JoyAI-Video-Edit (main) score 57

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:19.634548+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 2026-08-30T03:42:41.173939+00:00, confidence not recorded.
  - readme: https://github.com/jd-opensource/JoyAI-Video-Edit (fetched 2026-08-28T04:05:19.634548+00:00, sha bcd2626b46f0)
- Data as of 2026-08-30T08:39:29.467469+00:00.