We implemented the NAI picture style, which was mainly used in Korea, in the Anima model.
Although it can be used without trigger words, I believe that using the writer combination of trigger words will provide a more satisfying experience.

• base positive Prompt

• (masterpiece, best quality, score_9, year 2025, newest, highres, absurdres, very aesthetic, scenery, novel illustration, absurdres, masterpiece, score_9)

  artist/style prompt

• solo,

  safe,

• base negative Prompt

• (worst quality, low quality, early, old, score_1, score_2, score_3), easynegative, Hands-neg, lazyhand, lazyneg, negativeXL_D, cartoon, graphic, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured, long body, bad anatomy, bad hands, missing fingers, extra fingers, extra digits, fewer digits, cropped, very displeasing, (artist name:3), blurry, ((shiny skin)), (lowres, censor, (@taroimo \(00120014\):0.5), (@say hana:0.5), (@mx2j:0.5), (artist collaboration)), kusunoki midori, ((floating hair, pinup \(style\), higanbana no saku yoru ni):2)

• use Embedding

• https://civarchive.com/models/7808/easynegative?modelVersionId=9208

• https://civarchive.com/models/1065154/smooth-embeddings?modelVersionId=1668031

• https://civarchive.com/models/1302719/lazy-embeddings-for-all-illustrious-noobai-pony-sdxl-models-lazypositive-lazynegative-positive-and-negative-plus-more?modelVersionId=1860747

• https://civarchive.com/models/118418/negativexl?modelVersionId=134583

Diffusion in Low Bits : Automatic (fp16 LoRA) | Sampling Method : Euler a | Schedule Type : Normal | Sampling Steps : 25

Training has a cost!

If you want to support go to the Gallery (where the images are) and click " Add Post " to share your images and some blue buzz for generation/training! :)