SenseNova-U1.5 8B-MoT Native Unified Multimodal Model
September 9, 2026
SenseNova-U1.5 is an 8B Mixture-of-Tokens model that supports native 4K resolution without encoders or VAEs. It utilizes multi-expert on-policy distillation to handle specialized tasks like infographic generation, bilingual text rendering, and image editing.
HOW THIS AFFECTS YOU
●
builderYou can leverage an encoder-free architecture for higher fidelity visual reasoning and generation.
●
designerThe model's native support for 4K and specialized aesthetic experts enables advanced creative workflows.