EZBlender: Efficient 3D Editing with Plan-and-ReAct Agent

The official implementation of the paper EZBlender: Efficient 3D Editing with Plan-and-ReAct Agent

EZBlender is a lightweight, modular framework for automating 3D scene editing in Blender using Vision-Language Models (VLMs). It employs a multi-agent architecture where a central Planner decomposes high-level user prompts into specific sub-tasks for specialized agents (Modder, Material, Lighting, Camera, etc.).

Updates

• 4/10/2026: We are currently packaging the entire workflow into a skill for AI agents.
• 4/1/2026: We are currently working on the EZBench that benchmarking the ability of 3D editing among Models/Agents.
• 3/27/2026: We re-constructed the code using Codex CLI.
• 1/1/2026: Our paper is accepted by WACV2026@VALED.

Key Features

• Multi-Agent Pipeline: Intelligent task decomposition and execution.
• Vision-Feedback Loop: Autonomous refinement based on visual output analysis.
• Modular Design: Easily extendable with new agent roles or VLM providers.
• Automatic Debugging: Integrated error analysis and code fixing.

Architecture

1. Planner: Receives the prompt and current scene state, then creates a task list.
2. Specialized Agents:
   • Modder: Handles geometry and shape keys.
   • Material: Manages shaders and textures.
   • Lighting: Adjusts scene illumination.
   • Camera: Optimizes camera placement.
3. Refinement Loop: The Modder Critic evaluates results and iterates until satisfied.

Installation

1. Clone the repository:

   git clone https://github.com/your-username/EZBlender.git
   cd EZBlender

2. Install dependencies:

   pip install -r requirements.txt

3. Setup Environment Variables: Set your OpenAI credentials:

   /creds/openai.txt

Quick Start

Run the main workflow with a text prompt. Note that the assets are located in the blenderalch/ directory:

python main.py \
    --prompt "Make the object extremely bright glowing neon green" \
    --blender_exe "/path/to/blender" \
    --scene_blend "blenderalch/starter_blends/lotion.blend" \
    --init_script "blenderalch/blender_scripts/lighting_examples/lotion.py" \
    --render_script "blenderalch/blender_base/lighting_adjustments.py" \
    --out_dir "./output/test_run"

Acknowledgement

We sincerely appreciate the authors of BlenderAlchemy, as the BlenderAlchemy project provided a significant template/benchmark standard for 3D editing works. We have distilled the prompting logic from the original work into this lightweight, multi-agent framework, while preserving the original assets and research heritage within the blenderalch/ directory.

Contributing

Citation

If you find this project helpful to your research, please consider citing BibTeX:

@InProceedings{Wang_2026_WACV,
    author    = {Wang, Hao and Zhu, Wenhui and Tang, Shao and Wang, Zhipeng and Dong, Xuanzhao and Li, Xin and Chen, Xiwen and Bastola, Ashish and Huang, Xinhao and Wang, Yalin and Razi, Abolfazl},
    title     = {EZBlender: Efficient 3D Editing with Plan-and-ReAct Agent},
    booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops},
    month     = {March},
    year      = {2026},
    pages     = {1343-1352}
}