by agents-x-project · AI 工具 · ★ 153
PyVision: Agentic Vision with Dynamic Tooling 🎯Overview 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 , 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.
| Stars | 153 |
| Forks | 7 |
| Language | Python |
| Category | AI 工具 |
| Quality Score | 37.25/100 |
| Last Updated | 2025-07-22 |
| Created | 2025-06-27 |
| Platforms | python |
| Est. Tokens | ~604k |
These tools work well together with PyVision for enhanced workflows:
Explore other popular ai 工具 tools:
PyVision is [MTI-LLM@NeurIPS 2025] Official implementation of "PyVision: Agentic Vision with Dynamic Tooling.". It is categorized as a AI 工具 with 153 GitHub stars.
PyVision is primarily written in Python. It covers topics such as agent, computer-vision, mllm.
You can find installation instructions and usage details in the PyVision GitHub repository at github.com/agents-x-project/PyVision. The project has 153 stars and 7 forks, indicating an active community.