This is a workflow collection which uses the latest MiniMax H3  video model and lightx2v 4/8-step LoRA , plus a Comfyui UniblockSwap node to run on GPUs with 8GB of VRAM. It includes text to video , image to video, first and last frame and reference to video. You can use them all at one place.

Check  the tutorial for  detailed usage guide of the workflow:
https://youtu.be/2jv1I9-JULo