Train a LoRA that regenerates a masked region of a video while keeping the rest unchanged, blending the new content with its surroundings.
No supported capabilities recorded.
/v1/models/fal/fal-ai/ltx23-trainer-v2/inpaint