PointVLM​​, a novel multi-modal vision-language model that bridges 3D point cloud processing with natural language understanding to enable precise CAD model interpretation.