# bytedance/UNO

[ICCV 2025] 🔥🔥  UNO: A Universal Customization Method for Both Single and Multi-Subject Conditioning

Repository: https://github.com/bytedance/UNO
Canonical: https://ross.abutalabs.com/products/bytedance-uno
Homepage: https://bytedance.github.io/UNO/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: diffusion, diffusion-transformer, flux, image-generation, in-context-learning, subject-driven-generation, text-to-image, universal-image-generation
Last push: 2025-09-12T08:09:52+00:00

## Health v2 (maintenance only)
Score: 38/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 41, release rhythm 35, longevity 37
- inputs: {"age_days": 519, "days_push": 355, "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 1362, forks 79 (observed 2026-08-28T04:04:30.281555+00:00)

## What it is
UNO is a research framework from ByteDance for subject-driven image generation with diffusion transformers, supporting both single- and multi-subject customization. It builds on text-to-image models like FLUX using progressive cross-modal alignment and universal rotary position embedding (UnoPE).

## Use cases
- generate images of a specific subject from reference photos
- combine multiple subjects into one generated image
- customize a text-to-image model with my product or character
- create consistent character images across scenes
- subject-driven image generation with FLUX
- synthesize multi-subject training data for image generation

## When to choose
- you need high-fidelity single- or multi-subject image customization
- you want a research-grade diffusion transformer pipeline with pretrained checkpoints
- you are reproducing or extending the ICCV 2025 UNO paper

## When to avoid
- you need a polished end-user image editing app
- you lack a GPU or cannot run large diffusion models
- you need non-image modalities or lightweight inference on CPU

## Facets
- artifact type: library
- maturity: active
- function: image-processing, machine-learning, deep-learning, llm-training
- domain: image-processing, artificial-intelligence, deep-learning, computer-vision
- platform: python
- tags: diffusion-transformer, text-to-image, subject-driven-generation, flux, in-context-learning, image-customization, research-code, iccv-2025, gpu, linux

## Member repositories
- bytedance/UNO (main) score 38

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:30.281555+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 2026-08-30T04:41:29.313479+00:00, confidence not recorded.
  - readme: https://github.com/bytedance/UNO (fetched 2026-08-28T04:04:30.281555+00:00, sha ee679c80adbb)
  - homepage: https://bytedance.github.io/UNO/ (fetched 2026-08-29T11:59:02.382876+00:00, sha ea3dc5a22711)
- Data as of 2026-08-30T08:39:29.467469+00:00.