On August 21, 2026, SenseTime made a significant announcement: the official open-sourcing of its streamlined, native unified multimodal large model, SenseNova U1.5 Lite. This model has undergone meticulous optimization of its training data and post-training procedures, specifically tailored for real-world visual tasks. It boasts enhanced capabilities in following complex instructions, improving visual quality, refining text and layout, enabling native 4K output, facilitating image editing, and offering fine-grained visual control. As a result, it provides developers and creators with a more controllable, higher-definition, and cost-effective tool for visual creation. Presently, the model is available for open-source use worldwide, enabling developers and creators to deploy and explore it firsthand.