<div align="center"> <img src="assets/seedvr_logo.png" alt="SeedVR" width="400"/> </div>

SeedVR: Seeding Infinity in Diffusion Transformer Towards Generic Video Restoration

Jianyi Wang, Zhijie Lin, Meng Wei, Ceyuan Yang, Fei Xiao, Chen Change Loy, Lu Jiang

CVPR 2025 (Highlight)

<p> <a href="https://iceclear.github.io/projects/seedvr/"> <img src="https://img.shields.io/badge/SeedVR-Website-0A66C2?logo=safari&logoColor=white" alt="SeedVR Website" /> </a> <a href="https://huggingface.co/collections/ByteDance-Seed/seedvr-6849deeb461c4e425f3e6f9e"> <img src="https://img.shields.io/badge/SeedVR-Models-yellow?logo=huggingface&logoColor=yellow" alt="SeedVR Models" /> </a> <a href="https://huggingface.co/spaces/ByteDance-Seed/SeedVR2-3B"> <img src="https://img.shields.io/badge/SeedVR2-Space-orange?logo=huggingface&logoColor=yellow" alt="SeedVR2 Space" /> </a> <a href="https://github.com/numz/ComfyUI-SeedVR2_VideoUpscaler"> <img src="https://img.shields.io/badge/SeedVR-ComfyUI-blue?logo=googleplay&logoColor=blue" alt="SeedVR ComfyUI" /> </a> <a href="https://arxiv.org/abs/2501.01320"> <img src="https://img.shields.io/badge/SeedVR-Paper-red?logo=arxiv&logoColor=red" alt="SeedVR Paper on ArXiv" /> </a> <a href="https://www.youtube.com/watch?v=aPpBs_B2iCY" target='_blank'> <img src="https://img.shields.io/badge/Demo%20Video-%23FF0000.svg?logo=YouTube&logoColor=white" alt="SeedVR Video Demo on YouTube" /> </a> </p>

Why SeedVR: Conventional restoration models achieve inferior performance on both real-world and AIGC video restoration due to limited generation ability. Recent diffusion-based models improve the performance by introducing diffusion prior via ControlNet-like or adaptor-like architectures. Though gaining improvement, these methods generally suffer from constraints brought by the diffusion prior: these models suffer from the same bias as the prior, e.g., limited generation ability on small texts and faces, etc, and only work on fixed resolutions such as 512 or 1024. As a result, most of the existing diffusion-based restoration models rely on patch-based sampling, i.e., dividing the input video into overlapping spatial-temporal patches and fusing these patches using a Gaussian kernel at each diffusion step. The large overlap (e.g., 50% of the patch size), required for ensuring a coherent output without visible patch boundaries, often leads to considerably slow inference speed. This inefficiency becomes even more pronounced when processing long videos at high resolutions. SeedVR follows SOTA video generation training pipelines to tackle the key challenge in diffusion-based restoration, i.e., by enabling arbitrary-resolution restoration w/o relying on any pretrained diffusion prior and introducing advanced video generation technologies suitable for video restoration. Serving as the largest-ever diffusion transformer model towards generic video restoration, we hope SeedVR could push the frontiers of advanced VR and inspire future research in developing large vision models for real-world video restoration.

SeedVR2: One-Step Video Restoration via Diffusion Adversarial Post-Training

Jianyi Wang, Shanchuan Lin, Zhijie Lin, Yuxi Ren, Meng Wei, Zongsheng Yue, Shangchen Zhou, Hao Chen, Yang Zhao, Ceyuan Yang, Xuefeng Xiao, Chen Change Loy, Lu Jiang

ICLR 2026

<p> <a href="https://iceclear.github.io/projects/seedvr2/"> <img src="https://img.shields.io/badge/SeedVR2-Website-0A66C2?logo=safari&logoColor=white" alt="SeedVR Website" /> </a> <a href="https://huggingface.co/collections/ByteDance-Seed/seedvr-6849deeb461c4e425f3e6f9e"> <img src="https://img.shields.io/badge/SeedVR-Models-yellow?logo=huggingface&logoColor=yellow" alt="SeedVR Models" /> </a> <a href="https://huggingface.co/spaces/ByteDance-Seed/SeedVR2-3B"> <img src="https://img.shields.io/badge/SeedVR2-Space-orange?logo=huggingface&logoColor=yellow" alt="SeedVR2 Space" /> </a> <a href="https://github.com/numz/ComfyUI-SeedVR2_VideoUpscaler"> <img src="https://img.shields.io/badge/SeedVR-ComfyUI-blue?logo=googleplay&logoColor=blue" alt="SeedVR ComfyUI" /> </a> <a href="http://arxiv.org/abs/2506.05301"> <img src="https://img.shields.io/badge/SeedVR2-Paper-red?logo=arxiv&logoColor=red" alt="SeedVR2 Paper on ArXiv" /> </a> <a href="https://www.youtube.com/watch?v=tM8J-WhuAH0" target='_blank'> <img src="https://img.shields.io/badge/Demo%20Video-%23FF0000.svg?logo=YouTube&logoColor=white" alt="SeedVR2 Video Demo on YouTube" /> </a> </p>

Recent advances in diffusion-based video restoration (VR) demonstrate significant improvement in visual quality, yet yield a prohibitive computational cost during inference. While several distillation-based approaches have exhibited the potential of one-step image restoration, extending existing approaches to VR remains challenging and underexplored, due to the limited generation ability and poor temporal consistency, particularly when dealing with high-resolution video in real-world settings. In this work, we propose a one-step diffusion-based VR model, termed as SeedVR2, which performs adversarial VR training against real data. To handle the challenging high-resolution VR within a single step, we introduce several enhancements to both model architecture and training procedures. Specifically, an adaptive window attention mechanism is proposed, where the window size is dynamically adjusted to fit the output resolutions, avoiding window inconsistency observed under high-resolution VR using window attention with a predefined window size. To stabilize and improve the adversarial post-training towards VR, we further verify the effectiveness of a series of losses, including a proposed feature matching loss without significantly sacrificing training efficiency. Extensive experiments show that SeedVR2 can achieve comparable or even better performance compared with existing VR approaches in a single step.

<p align="center"><img src="assets/teaser.png" width="100%"></p>

📢 News

We sincerely thank all contributors from the open community for their valuable support.

• Jan. 27, 2026: SeedVR2 is accepted by ICLR2026.
• June 22, 2025: Huge thank NumZ, adrientoupet as well as the great community for building the ComfyUI !

<div align="center"> <a href="https://youtu.be/MBtWYXq_r60"> <img src="https://img.youtube.com/vi/MBtWYXq_r60/maxresdefault.jpg" width="50%" alt="SeedVR2 v2.5 Deep Dive Tutorial"> </a> </div>

• June 11, 2025: Repo created.

📮 Notice

Limitations: These are the prototype models and the performance may not perfectly align with the paper. Our methods are sometimes not robust to heavy degradations and very large motions, and shares some failure cases with existing methods, e.g., fail to fully remove the degradation or simply generate unpleasing details. Moreover, due to the strong generation ability, Our methods tend to overly generate details on inputs with very light degradations, e.g., 720p AIGC videos, leading to oversharpened results occasionally (especially on small resolutions, e.g., 480p).

🔥 Quick Start

1️⃣ Set up environment

git clone https://github.com/bytedance-seed/SeedVR.git
cd SeedVR
conda create -n seedvr python=3.10 -y
conda activate seedvr
pip install -r requirements.txt
pip install flash_attn==2.5.9.post1 --no-build-isolation

Install apex. If you encounter problems when installing apex from source, we provide two pre-build whl here:

• apex-0.1-cp39-cp39-linux_x86_64.whl: for python=3.9, torch=2.4.0, cuda=12.4
• apex-0.1-cp310-cp310-linux_x86_64.whl: for python=3.10, torch=2.4.0, cuda=12.1

pip install apex-0.1-cp310-cp310-linux_x86_64.whl

To use color fix, put the file color_fix.py to ./projects/video_diffusion_sr/color_fix.py.

2️⃣ Download pretrained checkpoint

# Take SeedVR2-3B as an example.
# See all models: https://huggingface.co/models?other=seedvr

from huggingface_hub import snapshot_download

save_dir = "ckpts/"
repo_id = "ByteDance-Seed/SeedVR2-3B"
cache_dir = save_dir + "/cache"

snapshot_download(cac