Skywork UniPic 3.0: Unified Multi-Image Composition via Sequence Modeling

TL;DR

Skywork UniPic 3.0 introduces a unified multi-image composition framework that leverages sequence modeling to enhance quality, efficiency, and flexibility in multi-image and single-image editing tasks, especially focusing on human-object interactions.

Contribution

The paper presents a novel sequence-modeling approach for multi-image composition, a comprehensive data pipeline, and a fast inference method, advancing the state-of-the-art in multi-image and single-image editing.

Findings

Achieves high-quality multi-image composition with only 700K training samples.

Produces high-fidelity samples in 8 steps, 12.5x faster than standard methods.

Outperforms existing models on multi-image composition benchmarks.

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.