BlenderRAG: High-Fidelity 3D Object Generation via Retrieval-Augmented Code Synthesis
For developers and designers needing to generate 3D objects from text, BlenderRAG offers a practical, fine-tuning-free improvement over existing LLM-based code generation.
BlenderRAG improves automatic Blender code generation from natural language by using retrieval-augmented generation on a curated dataset, boosting compilation success from 40.8% to 70.0% and semantic alignment from 0.41 to 0.77 CLIP similarity across four LLMs.
Automatic generation of executable Blender code from natural language remains challenging, with state-of-the-art LLMs producing frequent syntactic errors and geometrically inconsistent objects. We present BlenderRAG, a retrieval-augmented generation system that operates on a curated multimodal dataset of 500 expert-validated examples (text, code, image) across 50 object categories. By retrieving semantically similar examples during generation, BlenderRAG improves compilation success rates from 40.8% to 70.0% and semantic normalized alignment from 0.41 to 0.77 (CLIP similarity) across four state-of-the-art LLMs, without requiring fine-tuning or specialized hardware, making it immediately accessible for deployment. The dataset and code will be available at https://github.com/MaxRondelli/BlenderRAG.