# ByteDance-Seed/SeedVR

Repo for SeedVR2 (ICLR2026) & SeedVR (CVPR2025 Highlight)

Repository: https://github.com/ByteDance-Seed/SeedVR
Canonical: https://ross.abutalabs.com/products/seedvr
Homepage: https://iceclear.github.io/projects/seedvr2/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: diffusion-transformer, image-restoration, one-step-diffusion-model, video-restoration
Last push: 2026-01-27T06:02:24+00:00

## Health v2 (maintenance only)
Score: 47/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 64, release rhythm 35, longevity 32
- inputs: {"age_days": 449, "days_push": 218, "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 1334, forks 80 (observed 2026-08-28T04:04:25.201105+00:00)

## What it is
SeedVR/SeedVR2 are diffusion-transformer based models for generic real-world and AIGC video and image restoration, with SeedVR2 using adversarial post-training to achieve one-step high-resolution restoration. The repo provides model code, inference scripts, and pretrained weights.

## Use cases
- restore degraded real-world videos
- upscale and enhance AIGC-generated video to 1080p
- one-step fast video restoration
- restore small faces and text in low-quality video
- high-resolution video enhancement without patch artifacts

## When to choose
- you need state-of-the-art diffusion-based video restoration
- you want fast one-step inference at high resolutions
- you need arbitrary-resolution restoration without fixed 512/1024 limits

## When to avoid
- you lack a capable GPU
- you need lightweight real-time enhancement on CPU
- you only need simple classical denoising or deblocking

## Facets
- artifact type: library
- maturity: active
- function: video-processing, image-processing, machine-learning, deep-learning
- domain: computer-vision, image-processing, deep-learning, artificial-intelligence
- platform: python
- tags: diffusion-transformer, video-restoration, image-restoration, one-step-diffusion, video-upscaling, research-model, video, gpu, linux

## Member repositories
- ByteDance-Seed/SeedVR (main) score 47

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:25.201105+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 2026-08-30T04:44:27.106514+00:00, confidence not recorded.
  - readme: https://github.com/ByteDance-Seed/SeedVR (fetched 2026-08-28T04:04:25.201105+00:00, sha d0a9da1bbb9c)
  - homepage: https://iceclear.github.io/projects/seedvr2/ (fetched 2026-08-29T12:03:44.613572+00:00, sha a673a6ace648)
- Data as of 2026-08-30T08:39:29.467469+00:00.