This model is based on <a target="_blank" rel="ugc" href="https://civarchive.com/models/1188071/animagine-xl-40?modelVersionId=1337429">Animagine XL 4.0,</a> merged with my own model <a target="_blank" rel="ugc" href="https://civarchive.com/models/545481/ostandermix">OstAnderMix v1.0</a> and SDXL models including <a target="_blank" rel="ugc" href="https://civarchive.com/models/795765?modelVersionId=889818">Illustrious-XL</a> series.I tried to create a model that can generate 2.25D illustrations, while maintaining the ability to generate animated characters, which is a selling point of Animagine XL 4.0, as much as possible.<ul><li>License: <a rel="ugc" href="https://freedevproject.org/faipl-1.0-sd/">Fair AI Public License 1.0-SD</a></li><li>The merging process doesn't take into account the creation of NSFW works. The author personally does not recommend it.</li><li>Unauthorized reproduction of this model on other sites is prohibited.</li><li>If you would like to use my Checkpoint Online Generation, please click here.<ul><li><a target="_blank" rel="ugc" href="https://www.seaart.ai/ja/models/detail/cuicv55e878c73b7she0">SeaArt</a></li><li><a target="_blank" rel="ugc" href="https://pixai.art/model/1844771171397067380">Pixai</a></li></ul></li></ul>(Translated with <a target="_blank" rel="ugc" href="http://DeepL.com">DeepL.com</a>)<h3 id="recommended-setting-l354juaew">Recommended Setting</h3><ul><li>Steps: 25-30</li><li>CFG scale: 5-7</li><li>Sampler : Euler a</li><li>Scheduler : Normal, Simple, SGM Uniform</li><li>Use ADetailer as needed</li></ul><h3 id="positive-prompt-3cd2cmtyk">Positive Prompt</h3><pre><code>masterpiece, high score, great score, absurdres</code></pre><h3 id="negative-prompt-gw0trcsax">Negative Prompt</h3><pre><code>lowres, bad anatomy, bad hands, text, error, missing finger, extra digits, fewer digits, cropped, worst quality, low quality, low score, bad score, average score, signature, watermark, username, blurry</code></pre>

OstAnderAni4MIX

Ver.1.3 (2025/03/10)This model is based on Animagine XL v4.0 Opt and was merged back from scratch. When I created the merged model, I also tried to reduce noise and depict the eyes, but I wanted to preserve as much of the v1.0 features as possible, so I am not sure how far I was able to achieve this in v1.3. Thank you to the creators of each of the models used in this model merge.(Translated by DeepL.com)このモデルはAnimagine XL v4.0 Optをベースに1からマージしなおしています。今回は、v1.0で気になった眼の描写・ノイズの低減も図りましたが、v1.0の特徴もできるだけ残したかったので、v1.3でどこまで実現できたかはわかりません。 またこのモデルのマージに使用した各モデルの作成者の皆様、いつもありがとうございます。<ol><li>(XXMix_9realisticSDXL v1.0 x 0.5 +fuduki_Mix v2.0 x 0.5) x 0.8 + 9527 Detail Realistic XL v7.0-Mark2 x 0.2</li><li>OstAnderMix v1.0 + (Animagine XL 4.0 v4Opt - Animagine XL V3.1 v3.1) x 1 (trainDifference)</li><li>(2) + (1) (GRAD_V)</li><li>Animagine XL 4.0 v4Opt x 0.05 + (2) x 0.95 (cosineB)</li><li>OstAnderAni4MIX v1.0 + (Animagine XL 4.0 v4Opt - Animagine XL 4.0 v4.0) x 1_trainDifference)x0.05</li><li>(5) x 0.05 + (3) x 0.95 (cosineB)</li><li>Animagine XL 4.0 v4Opt+ (uncannyvalley noob 3d V1 D - NoobAI-XL (NAI-XL) epsilonPred 1.1-Version) x 0.65</li><li>(7) + (copycat-noob MR_v1.0(v-pred) - NoobAI-XL (NAI-XL) V-Pred 1.0-Version) x 0.45</li><li>(6) + (8) (SMOOTHSTEP)</li><li>(8) + (Rillusm v2.0 - illustrious-XL v0.1) x 0.5</li><li>(4) x 0.6 + (10) x 0.4 (cosineB)</li><li>(9) + (11) (FLAT_25)</li><li>(12) + OstAnderAni4MIX v1.0 (GRAD_V)</li></ol>

Files with SHA256: b327490153d2f447bdb75db8d147dbea08e66bd9fc8ebd13982fa498491c8ef0