Visual content creation tasks demand a nuanced understanding of design conventions and creative workflows-capabilities challenging for general models, while workflow-based agents lack specialized knowledge for autonomous creative planning. To overcome these challenges, we propose VisionCreator, a native visual-generation agentic model that unifies Understanding, Thinking, Planning, and Creation (UTPC) capabilities within an end-to-end learnable framework. Our work introduces four key contributions: (i) VisGenData-4k and its construction methodology using metacognition-based VisionAgent to generate high-quality creation trajectories with explicit UTPC structures; (ii) The VisionCreator agentic model, optimized through Progressive Specialization Training (PST) and Virtual Reinforcement Learning (VRL) within a high-fidelity simulated environment, enabling stable and efficient acquisition of UTPC capabilities for complex creation tasks; (iii) VisGenBench, a comprehensive benchmark featuring 1.2k test samples across diverse scenarios for standardized evaluation of multi-step visual creation capabilities; (iv) Remarkably, our VisionCreator-8B/32B models demonstrate superior performance over larger closed-source models across multiple evaluation dimensions. Overall, this work provides a foundation for future research in visual-generation agentic systems.
@article{arxiv.2603.02681,
title = {VisionCreator: A Native Visual-Generation Agentic Model with Understanding, Thinking, Planning and Creation},
author = {Jinxiang Lai and Zexin Lu and Jiajun He and Rongwei Quan and Wenzhe Zhao and Qinyu Yang and Qi Chen and Qin Lin and Chuyue Li and Tao Gao and Yuhao Shan and Shuai Shao and Song Guo and Qinglin Lu},
journal= {arXiv preprint arXiv:2603.02681},
year = {2026}
}