VisionCreator: A Native Visual-Generation Agentic Model with Understanding, Thinking, Planning and Creation

TL;DR

VisionCreator is a unified, end-to-end visual-generation model that integrates understanding, thinking, planning, and creation capabilities, enabling autonomous complex visual content creation.

Contribution

It introduces a novel framework with specialized data, training methods, and benchmarks for developing agentic models with multi-step visual creation skills.

Findings

VisionCreator-8B/32B outperform larger closed-source models.

High-quality creation trajectories generated with explicit UTPC structures.

Effective training via Progressive Specialization and Virtual Reinforcement Learning.

Abstract

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.