# JiuhaiChen/BLIP3o

Official implementation of BLIP3o-Series

Repository: https://github.com/JiuhaiChen/BLIP3o
Canonical: https://ross.abutalabs.com/products/blip3o
Language: Python
License Family: other
Last push: 2025-11-29T22:25:14+00:00

## Health v2 (maintenance only)
Score: 44/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 54, release rhythm 35, longevity 35
- inputs: {"age_days": 497, "days_push": 277, "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 1666, forks 79 (observed 2026-08-28T04:05:19.565753+00:00)

## What it is
Official implementation of the BLIP3o-Series, a unified autoregressive-plus-diffusion model for text-to-image generation and editing. It combines an autoregressive model that produces intermediate features with a diffusion model for image synthesis, trained with discrete image token supervision and GRPO reinforcement learning.

## Use cases
- generate images from text prompts
- train a unified autoregressive diffusion image generation model
- fine-tune a text-to-image model with instruction tuning data
- apply GRPO reinforcement learning to improve prompt alignment in image generation
- improve text rendering in generated images
- reproduce BLIP3o-NEXT research results
- pretrain a multimodal image generation model on open caption datasets

## When to choose
- you need an open-source, fully reproducible text-to-image model with training code and data
- you want to experiment with combining autoregressive and diffusion architectures
- you want to apply RLHF-style GRPO training to image generation models
- you are researching discrete image token supervision for multimodal models

## When to avoid
- you just want a plug-and-play image generator without training or research work
- you lack GPU/Slurm infrastructure for large model training
- you need a commercially licensed model (no license is specified)
- you need a stable production-ready library rather than research code

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, llm-training, image-processing
- domain: artificial-intelligence, deep-learning, image-processing, large-language-models
- platform: python
- tags: text-to-image, autoregressive-diffusion, image-generation, grpo, reinforcement-learning, research-code, multimodal, linux, gpu

## Member repositories
- JiuhaiChen/BLIP3o (main) score 44

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:19.565753+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 2026-08-30T03:42:56.928977+00:00, confidence not recorded.
  - readme: https://github.com/JiuhaiChen/BLIP3o (fetched 2026-08-28T04:05:19.565753+00:00, sha 249f703be707)
- Data as of 2026-08-30T08:39:29.467469+00:00.