On August 25, during its inaugural Technology Open Day (TechDay), Qunhe Technology showcased the latest advancements in its space large model: the state-of-the-art SpatialLM 1.5 spatial language model and the SpatialGen spatial generation model. The company also announced its open source strategy for these new models, marking a pioneering move in the industry. Designed specifically for the recognition and generation of 3D indoor scenes, this large model system boasts remarkable capabilities in realistic holographic navigation, structured interactivity, and complex indoor scene processing. SpatialLM 1.5 enables the creation of structured 3D scenes through a conversational interface, while SpatialGen can generate multi-view images and 3D scenes based on textual descriptions and 3D layouts. These innovations are poised to propel the growth of spatial intelligence technology and will gradually become available to developers worldwide on platforms like HuggingFace and GitHub, underscoring Qunhe's commitment to open collaboration.