Train a LoRA that expands the video frame outward, keeping an inner rectangle fixed and generating the surrounding region.
No supported capabilities recorded.
/v1/models/fal/fal-ai/ltx23-trainer-v2/outpaint