Paper · arXiv 2507.07998
PyVision: Agentic Vision with Dynamic Tooling
Shitian Zhao, Haoquan Zhang, Shaoheng Lin, Ming Li, Qilong Wu, Kaipeng Zhang, Chen Wei
PyVision enables MLLMs to autonomously generate, execute, and refine Python-based tools for visual reasoning, achieving significant performance improvements across benchmarks.
Abstract
LLMs are increasingly deployed as agents, systems capable of planning, reasoning, and dynamically calling external tools. However, in visual reasoning, prior approaches largely remain limited by predefined workflows and static toolsets. In this report, we present PyVision, an interactive, multi-turn framework that enables MLLMs to autonomously generate, execute, and refine Python-based tools tailored to the task at hand, unlocking flexible and interpretable problem-solving. We develop a taxonomy of the tools created by PyVision and analyze their usage across a diverse set of benchmarks. Quantitatively, PyVision achieves consistent performance gains, boosting GPT-4.1 by +7.8% on V* and Claude-4.0-Sonnet by +31.1% on VLMsAreBlind-mini. These results point to a broader shift: dynamic tooling allows models not just to use tools, but to invent them, advancing toward more agentic visual reasoning.