# Saiyan-World/goku

[CVPR2025 Highlight] Video Generation Foundation Models: https://saiyan-world.github.io/goku/

Repository: https://github.com/Saiyan-World/goku
Canonical: https://ross.abutalabs.com/products/goku
Homepage: https://saiyan-world.github.io/goku/
Language: Python
License Family: other
Last push: 2025-02-19T14:33:19+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 7, release rhythm 35, longevity 40
- inputs: {"age_days": 570, "days_push": 560, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2905, forks 308 (observed 2026-08-28T04:07:29.055828+00:00)

## What it is
Goku is a family of flow-based (rectified flow Transformer) foundation models for joint image and video generation, released by HKU and ByteDance. The repository provides model code, weights, and inference pipelines for text-to-video, image-to-video, and text-to-image generation.

## Use cases
- generate videos from text prompts
- animate a still image into a video
- generate high-quality images from text descriptions
- research rectified flow transformer models for visual generation
- benchmark video generation models on VBench or MovieGenBench

## When to choose
- you need state-of-the-art open text-to-video or image-to-video generation
- you are researching flow-based generative models and want strong baselines
- you have GPU resources for running large generative models locally

## When to avoid
- you need a lightweight or CPU-only image/video generator
- you require a permissive license for commercial use (no license is specified)
- you only need simple video editing rather than generative synthesis

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, video-processing, image-processing, llm-inference
- domain: deep-learning, image-processing, artificial-intelligence
- platform: python
- tags: text-to-video, image-to-video, text-to-image, rectified-flow, diffusion-transformer, cvpr2025, foundation-model, generative-ai, video, gpu, linux

## Member repositories
- Saiyan-World/goku (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:29.055828+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 2026-08-30T07:34:30.242186+00:00, confidence not recorded.
  - readme: https://github.com/Saiyan-World/goku (fetched 2026-08-28T04:07:29.055828+00:00, sha 0ac04934f93f)
  - homepage: https://saiyan-world.github.io/goku/ (fetched 2026-08-29T09:49:42.723135+00:00, sha 8120dcd4a5c9)
- Data as of 2026-08-30T08:39:29.467469+00:00.