Pony Diffusion is a community-driven text-to-image diffusion model built on Stable Diffusion, known for high-quality stylized and non-photorealistic image generation. V6 XL trains on ~2.6M images with score-tag prompting (score_9/score_8_up) and broad Danbooru/CLIP caption support. V7 expands to ~10M images with style grouping and better SFW coverage. Free to try online via Hugging Face Spaces; no sign-up required.
