SenseTime Launches SenseNova U1 Pro Image Model with 8K Support
SenseTime has officially released the SenseNova U1 Pro image creation model, integrating it into the Xiaohuanxiong platform and the SenseNova API service. The model supports ultra-high resolution outputs up to 8K and custom aspect ratios, using built-in interleaved vision-text Chain-of-Thought reasoning. This release advances AI image generation toward professional design workflows, enabling accurate text rendering, structured infographics, and detailed architectural renderings. It demonstrates how Chinese AI developers are pushing commercial multimodal capabilities to rival leading global image models. SenseNova U1 Pro emphasizes production-ready visual quality, enabling accurate multi-module layout design and clear Chinese/English typography within generated images. However, the model is available only via proprietary APIs and commercial applications without open-source weight access.
## BACKGROUND
SenseNova is SenseTime's family of foundational AI models that powers their enterprise and consumer multimodal services. Interleaved vision-text Chain-of-Thought is an advanced multimodal reasoning technique where the model alternates between visual feature processing and textual reasoning steps to handle complex, spatially aware layout tasks.