JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion

Hugging Face Daily Papers Papers
video-editing diffusion autoregressive real-time distillation

Summary

JoyAI-Video-Edit is a 16B-parameter autoregressive diffusion framework for real-time, open-ended video editing, achieving 720p editing at ~30 FPS on a single NVIDIA B200 GPU.

Real-time video editing requires low-latency causal generation with bounded computational resources while preserving source fidelity and long-term temporal consistency. We present JoyAI-Video-Edit, a 16B-parameter autoregressive diffusion framework for real-time, open-ended video editing without access to future frames or a predefined video duration. Our method combines chunk-wise autoregressive adaptation, Source-Anchored Distribution Matching Distillation (SA-DMD), and Long-Horizon Autoregressive Distillation to reduce train--inference mismatch, preserve source fidelity during two-step generation, and mitigate accumulated temporal drift. Extensive automatic and human evaluations show that JoyAI-Video-Edit substantially outperforms existing streaming editors and remains competitive with strong offline systems on both short and long videos. The complete system achieves end-to-end 720p video editing at approximately 30 FPS on a single Nvidia B200 GPU. Code is available at https://github.com/jd-opensource/JoyAI-Video-Edit.
Original Article
View Cached Full Text

Cached at: 08/05/26, 05:43 AM

Paper page - JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion

Source: https://huggingface.co/papers/2608.03974 Published on Aug 4

#1 Paper of the day Authors:

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

Real-timevideoeditingrequireslow-latencycausalgenerationwithboundedcomputationalresourceswhilepreservingsourcefidelityandlong-termtemporalconsistency.WepresentJoyAI-Video-Edit,a16B-parameterautoregressivediffusionframeworkforreal-time,open-endedvideoeditingwithoutaccesstofutureframesorapredefinedvideoduration.Ourmethodcombineschunk-wiseautoregressiveadaptation,Source-AnchoredDistributionMatchingDistillation(SA-DMD),andLong-HorizonAutoregressiveDistillationtoreducetrain--inferencemismatch,preservesourcefidelityduringtwo-stepgeneration,andmitigateaccumulatedtemporaldrift.ExtensiveautomaticandhumanevaluationsshowthatJoyAI-Video-Editsubstantiallyoutperformsexistingstreamingeditorsandremainscompetitivewithstrongofflinesystemsonbothshortandlongvideos.Thecompletesystemachievesend-to-end720pvideoeditingatapproximately30FPSonasingleNvidiaB200GPU.Codeisavailableathttps://github.com/jd-opensource/JoyAI-Video-Edit.

View arXiv pageView PDFGitHub40Add to collection

Get this paper in your agent:

hf papers read 2608\.03974

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper1

#### jdopensource/JoyAI-Video-Edit Video-to-Video• Updatedabout 2 hours ago • 11

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2608.03974 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2608.03974 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

jdopensource/JoyAI-Echo

Hugging Face Models Trending

JD Open Source releases JoyAI-Echo (Echo-LongVideo), a text-to-audio-video diffusion model capable of generating minute-level multi-shot videos with consistent character identity and voice, using DMD distillation for 7.5x speedup.