Animagine XL

Animagine XL is an anime-focused Stable Diffusion XL model series by Cagliostro Research Lab. The current main line is 4.0 (Anim4gine), rebuilt from SDXL 1.0 with approximately 8.4 million anime-style images and about 2,650 GPU hours. It delivers more artistic and prompt-following results than 3.1, with extra quality, score, year, and rating controls. Excels at anime characters, detailed illustrations, and recent series knowledge. Free to try via browser playgrounds on the website. Animagine XL excels at anime characters, detailed illustrations, and recent series knowledge. Recommended settings include Euler a sampler, 25-28 steps, CFG scale 4-7 (often 5), and standard SDXL resolutions. Early 3.x versions refined hands and knowledge. The model is publicly available under the CreativeML Open RAIL++-M license. Users can try it for free via browser playgrounds on the website. The model supports various aspect ratios and can generate anything from simple character portraits to complex multi-character scenes with detailed backgrounds. It includes built-in quality tags that let users control the output quality level, from sketch to masterpiece. The rating system allows filtering by content rating, and the year tags help maintain consistency with specific anime art styles from different periods. The Animagine XL series is continuously updated with community feedback, ensuring it stays at the forefront of anime AI generation technology. Users can join the Discord community to share creations and get support.