Try flux lora by Inference API
Generate images from text prompts
text-to-image
Apply the motion of a video on a portrait
Lots of Liberal ImgGen Models