Multimodal Ai

Ling-3.2-flash adds image and video input without changing its headline parameter count

Ant’s new API model keeps the 560B total, 25B active and million-token context figures. Its supported inputs are the consequential difference.