Pony Diffusion V6 XL Reviews

It understands prompts much better than any SD 1.5 or Illustrious model I have tested. It isn't good at even halfway complicated prose, but I don't feel like guessing the exact spelling and format of a tag (cross-eyed/cross eyed/crossed eyes/eyes crossed) is as important to producing the intended result. I don't need LoRAs to do nearly as much either, such as achieve certain facial expressions or specific poses.