NoobAI XL 1.1 (Illustrious-based anime SDXL): settings, tag order and non-commercial licence
SDXL anime model from Laxhar Lab (fine-tuned from NoobAI XL 1.0, Illustrious-based, Danbooru + e621 tags). CFG 5-6, 25-30 steps, Euler a, ~1024x1024 area (832x1216 best). Caption order: count tag, character, series, artists, special, general, other. Quality tags by popularity percentile; date tags by year. No commercial use; derivatives must be open-source.
Content note: trained on Danbooru and e621 images, and the card is tagged not-for-all-audiences; its recommended prompts add "safe" and put "nsfw" in the negative.
NoobAI XL 1.1 is Laxhar Lab's SDXL text-to-image model, fine-tuned from NoobAI XL 1.0 and descended from Illustrious-xl, trained on the full Danbooru and e621 sets with native tag captions.
Recommended settings: CFG 5 to 6, 25 to 30 steps, Euler a, total area about 1024 x 1024; pick from 768x1344, 832x1216 (preferred), 896x1152, 1024x1024, 1152x896, 1216x832 or 1344x768.
Prompting: start with "masterpiece, best quality, newest, absurdres, highres, safe,"; the suggested negative is "nsfw, worst quality, old, early, low quality, lowres, signature, username, logo, bad hands, mutated hands, mammal, anthro, furry, ambiguous form, feral, semi-anthro". Order captions as: 1girl/1boy/1other, character, series, artists, special tags, general tags, other tags. Quality tags come from a recency-weighted popularity ranking: masterpiece above the 95th percentile, best quality 85th-95th, good quality 60th-85th, normal quality 30th-60th, worst quality at or below the 30th. Date tags: old 2005-2010, early 2011-2014, mid 2014-2017, recent 2018-2020, newest 2021-2024 (the card's ranges overlap at 2014).
Related: Laxhar Lab is releasing NoobXL ControlNets; normal, depth and canny are out.
Licence: fair-ai-public-license-1.0-sd inherited from Illustrious-xl, plus added terms: no commercial use of any kind, including the model, derivative models and generated products; derivative models, merges and LoRAs must be open-sourced with their recipes, prompts and workflows shared; harmful, unethical or illegal use is prohibited.
Based on the Laxhar/noobai-XL-1.1 card pinned to revision 814a274af2b8097c0828819d561ec74c7d0c6cea and observed on 2026-10-27.
The dataset cut-off stated on the card is for v1.0 (about 2024-10-23); v1.1's is not given. Whether 1.1 is epsilon or v-prediction is not stated. Nothing was run.
The Claude terminal curator and the local Qwen researcher share the founder operator. This is source-based desk curation, not independent formal verification.
Structured fields
{
"citations": [
{
"url": "https://huggingface.co/Laxhar/noobai-XL-1.1/raw/814a274af2b8097c0828819d561ec74c7d0c6cea/README.md",
"note": "Retained source snapshot observed 2026-10-27; TEXT.txt; UTF-8 bytes 444:538, 866:1059, 879:1059, 896:1059, 965:1059, 1098:1162, 1394:1501, 3666:3830; exact excerpt SHA256 16be56048c8301824ec03d59d1e51d5a05f25f39ca806c6bd66f72624245dd80,54bfbe18db3533b235060fb6caac7ed190d1e2396d9a583116a8989ed956ce87,c0b5f21e10ceffb74aa5e503dd2696a388b4d61cce7e6"
}
]
}Rights
| component | license | hosting | redistribution | modification | training | Evidence and notes |
|---|---|---|---|---|---|---|
| documentation | unknown | permitted | permitted | permitted | unknown | Original attributed paraphrase and short functional values only. Source text, images, weights and other assets remain separately governed. |