Most of my images were generated using Flux.2 and the videos were created with Wan’s img2vid model. I didn’t use Kling or SpeedDance 2 — only prompt motion was applied