Skywork UniPic 3.0: Unified Multi-Image Composition via Sequence Modeling
Hongyang Wei, Hongbo Liu, Zidong Wang, Yi Peng, Baixin Xu, Size Wu, Xuying Zhang, Xianglong He, Zexiang Liu, Peiyu Wang, Xuchen Song, Yangguang Li, Yang Liu, Yahui Zhou
TL;DR
Skywork UniPic 3.0 presents a unified sequence-modeling framework that jointly handles single-image editing and multi-image HOI-centric composition across 1–6 inputs and variable resolutions within a 1024x1024 budget. It introduces a carefully curated 215K HOI-focused dataset, a latent patch-based sequence representation, and a post-training pipeline that combines trajectory mapping and distribution matching to achieve high-quality results in just 8 inference steps. The approach yields state-of-the-art performance on standard editing benchmarks and surpasses strong baselines like Nano-Banana and Seedream 4.0 on a new MultiCom-Bench for multi-image composition, validating the data pipeline and unified paradigm. The work demonstrates practical impact by enabling flexible, efficient, and interactive generative capabilities that generalize across editing and composition tasks, with public release of code, models, and data.
Abstract
The recent surge in popularity of Nano-Banana and Seedream 4.0 underscores the community's strong interest in multi-image composition tasks. Compared to single-image editing, multi-image composition presents significantly greater challenges in terms of consistency and quality, yet existing models have not disclosed specific methodological details for achieving high-quality fusion. Through statistical analysis, we identify Human-Object Interaction (HOI) as the most sought-after category by the community. We therefore systematically analyze and implement a state-of-the-art solution for multi-image composition with a primary focus on HOI-centric tasks. We present Skywork UniPic 3.0, a unified multimodal framework that integrates single-image editing and multi-image composition. Our model supports an arbitrary (1~6) number and resolution of input images, as well as arbitrary output resolutions (within a total pixel budget of 1024x1024). To address the challenges of multi-image composition, we design a comprehensive data collection, filtering, and synthesis pipeline, achieving strong performance with only 700K high-quality training samples. Furthermore, we introduce a novel training paradigm that formulates multi-image composition as a sequence-modeling problem, transforming conditional generation into unified sequence synthesis. To accelerate inference, we integrate trajectory mapping and distribution matching into the post-training stage, enabling the model to produce high-fidelity samples in just 8 steps and achieve a 12.5x speedup over standard synthesis sampling. Skywork UniPic 3.0 achieves state-of-the-art performance on single-image editing benchmark and surpasses both Nano-Banana and Seedream 4.0 on multi-image composition benchmark, thereby validating the effectiveness of our data pipeline and training paradigm. Code, models and dataset are publicly available.