SenseNova-U1 thinks in images > Not "text in – image out." It generates images natively, mid-reasoning, as part of the thinking chain. > Input question – Interleaved text + Image reasoning – Structured output.

→ View original post on X — @testingcatalog