# SHI-Labs/Versatile-Diffusion

Versatile Diffusion: Text, Images and Variations All in One Diffusion Model, arXiv 2022 / ICCV 2023

Repository: https://github.com/SHI-Labs/Versatile-Diffusion
Canonical: https://ross.abutalabs.com/products/versatile-diffusion
Homepage: https://arxiv.org/abs/2211.08332
Language: Python
License: MIT
License Family: permissive
Last push: 2023-08-10T10:17:01+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1400, "days_push": 1119, "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 84 (observed 2026-08-28T04:04:24.884052+00:00)

## What it is
Versatile Diffusion is the official PyTorch implementation of a unified multi-flow multimodal diffusion model that handles text-to-image, image-to-text, image-variation, and text-variation in one model. It supports extensions like style-semantics disentanglement, dual-context blending, and latent image editing, with a WebUI and Hugging Face demo.

## Use cases
- generate images from text prompts
- generate text captions describing an image
- create variations of an existing image
- blend text and image guidance for generation
- disentangle style and semantics in generated images
- edit images via latent image-to-text-to-image editing

## When to choose
- you want one unified model covering multiple text-image generation tasks
- you need research-grade multimodal diffusion with swappable flows
- you want to experiment with dual-guided or disentanglement generation

## When to avoid
- you need the latest state-of-the-art image quality from newer diffusion models
- you need video, audio, or 3D generation, which is not yet supported
- you want a lightweight model without GPU resources

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, image-processing, nlp
- domain: artificial-intelligence, deep-learning, image-processing
- platform: python
- tags: diffusion-models, text-to-image, image-to-text, multimodal, generative-ai, stable-diffusion, pytorch, iccv-2023, natural-language-processing, gpu

## Member repositories
- SHI-Labs/Versatile-Diffusion (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:24.884052+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 2026-08-30T04:44:32.076256+00:00, confidence not recorded.
  - readme: https://github.com/SHI-Labs/Versatile-Diffusion (fetched 2026-08-28T04:04:24.884052+00:00, sha 41802b8a3910)
  - homepage: https://arxiv.org/abs/2211.08332 (fetched 2026-08-29T12:03:50.923229+00:00, sha 3bffde17aed1)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T12:03:50.932165+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T12:03:50.935268+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T12:03:50.937023+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T12:03:50.933789+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.