I'm excited to share Illustrious XL with you all today — the open-source AI illustration model that's been quietly powering thousands of derivative models across the community. The problem we set out to solve: Standard SDXL models are capped at 1024×1024 native resolution, forcing creators to rely on upscalers that hallucinate details and blur fine lines. If you've ever tried generating anime art and ended up with weird hands or distorted faces, you know the pain. What makes Illustrious XL different? high resolution — no upscalers needed, just crisp lineart and rich details straight out of the box Hybrid prompting — write in plain English and append Danbooru tags. Both just work. No more fighting with your prompt syntax. Fully open-source — released under CreativeML OpenRAIL-M / Fair AI Public License. Fine-tune it, build LoRAs on it, make it yours. ⚡ Compatibility — drop it into ComfyUI, A1111, Forge, or Draw Things. Zero friction. We've iterated aggressively from v0.1 all the way to v3.5 VPred, pushing SDXL to its absolute technical limits. And now we're transitioning to the Lumina architecture — so this is just the beginning.

Illustrious XL

Illustrious XL - High-Resolution Anime-Style AI Generation

Illustrious XL - High-Resolution Anime-Style AI Generation