If you want to support me, bidding buzz on the weekly <a target="_blank" rel="ugc" href="https://civarchive.com/auctions">auctions page</a> for the model helps me keep it up for on-site generation. <h1 id="anima-base-v1">Anima Base v1 </h1>Text encoder and VAE can be found on this <a target="_blank" rel="ugc" href="https://huggingface.co/circlestone-labs/Anima">HuggingFace page</a>.<h2 id="recommended-settings-:">Recommended settings :</h2><ul><li>Positive prompt : masterpiece, best quality</li><li>Negative prompt : worst quality, low quality, jpeg artifacts, signature, sepia, fewer digits, extra digits, bad hands, bad anatomy, watermark, censored, score_1, score_2, score_3</li><li>Sampler : ER SDE / Euler a - 24 steps</li><li>CFG Scale : 4-5 </li></ul>I highly recommend using the <a target="_blank" rel="ugc" href="https://github.com/Bing-su/adetailer">ADetailer</a> extension for better faces and eyes. The Anima iteration of this checkpoint is a bit of a departure from the Illustrious v2 version, going for a slightly less flat style overall, but keeping a colorful, sharp 2D anime style for faces while getting a little painterly side to bodies and backgrounds, as well as shading. Just like with other Anima based checkpoints, artist styles work, and natural language to enhance your prompts or add text also works.Thanks to @ComradeAnanas, @Manityro, and @justTNP for helping create this checkpoint.<h1 id="illustriousxlnoobai-eps1.1">IllustriousXL/NoobAI EPS1.1</h1><h2 id="recommended-settings-:">Recommended settings :</h2><ul><li>Positive prompt : masterpiece, best quality</li><li>Negative prompt : worst quality, low quality, bad anatomy, watermark, censored</li><li>Sampler : Euler a</li><li>CFG Scale : 5-7</li></ul>I highly recommend using the <a target="_blank" rel="ugc" href="https://github.com/Bing-su/adetailer">ADetailer</a> extension for better faces and eyes.For v2, I merged v1 with some NoobAI models, @GarejiSama's <a target="_blank" rel="ugc" href="https://civarchive.com/models/1649312?modelVersionId=1866796">AdAstra</a> and @Manityro's <a target="_blank" rel="ugc" href="https://civarchive.com/models/1732765/vermilion">Vermilion</a>, making v2 much more compatible with NoobAI based LoRAs. The baked in VAE is Anzhc's, which you can get <a target="_blank" rel="ugc" href="https://huggingface.co/Anzhc/Anzhcs-VAEs/blob/main/SDXL%20Anime%20VAE%20Dec-only%20B3.safetensors">here</a>, gives a much sharper look to images, reduced noise, and makes colors less bland overall. LustrousMix v2 generations are more colorful than v1's, with more contrast and sharpness. There is a slight art style change as well, with more mature faces. <img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/babdaa02-2197-4a29-9b37-13d2467db4e2/width=525/babdaa02-2197-4a29-9b37-13d2467db4e2.jpeg" /><h3 id="hope-you-enjoy!">Hope you enjoy!</h3>

Hoseki - LustrousMix [Anima v1.0]

hosekiLustrousmixAnima_illustNoobaiEPS11V2.safetensors

Hoseki - LustrousMix [IllustriousXL/NoobAI EPS1.1]

Illust/NoobAI EPS1.1 - v2

Anima v1.0

Illustrious v1

creator: HosekiAnima Base v1Recommended settings :<ul><li>Positive prompt : masterpiece, best quality</li><li>Negative prompt : worst quality, low quality, bad anatomy, jpeg artifacts, signature, sepia, fewer digits, extra digits, bad hands, bad anatomy, watermark, censored, score_1, score_2, score_3</li><li>Sampler : ER SDE / Euler a - 24 steps</li><li>CFG Scale : 4-5</li></ul>I highly recommend using the <a target="_blank" rel="ugc" href="https://github.com/Bing-su/adetailer">ADetailer</a> extension for better faces and eyes. The Anima iteration of this checkpoint is a bit of a departure from the Illustrious v2, going for a slightly less flat style overall, but keeping a colorful, sharp 2D anime style for faces while getting a little painterly side to bodies and backgrounds, as well as shading. Just like with other Anima based checkpoints, artist styles work, and natural language to enhance your prompts or add text also works.Thanks to @ComradeAnanas, @Manityro, and @justTNP for helping create this checkpoint.IllustriousXL/NoobAI EPS1.1Recommended settings :<ul><li>Positive prompt : masterpiece, best quality</li><li>Negative prompt : worst quality, low quality, bad anatomy, watermark, censored</li><li>Sampler : Euler a</li><li>CFG Scale : 5-7</li></ul>I highly recommend using the <a target="_blank" rel="ugc" href="https://github.com/Bing-su/adetailer">ADetailer</a> extension for better faces and eyes.For v2, I merged v1 with some NoobAI models, @GarejiSama's <a target="_blank" rel="ugc" href="https://civitai.com/models/1649312?modelVersionId=1866796">AdAstra</a> and @Manityro's <a target="_blank" rel="ugc" href="https://civitai.com/models/1732765/vermilion">Vermilion</a>, making v2 much more compatible with NoobAI based LoRAs. The baked in VAE is Anzhc's, which you can get <a target="_blank" rel="ugc" href="https://huggingface.co/Anzhc/Anzhcs-VAEs/blob/main/SDXL%20Anime%20VAE%20Dec-only%20B3.safetensors">here</a>, gives a much sharper look to images, reduced noise, and makes colors less bland overall. LustrousMix v2 generations are more colorful than v1's, with more contrast and sharpness. There is a slight art style change as well, with more mature faces.Hope you enjoy!

Hoseki - LustrousMix [IllustriousXL]

Hoseki - LustrousMix [Anima]

NoobAI EPS1.1-v2

Recommended settings :<ul><li>Positive prompt : masterpiece, best quality</li><li>Negative prompt : worst quality, low quality, bad anatomy, watermark, censored</li><li>Sampler : Euler a</li><li>CFG Scale : 5-7</li></ul>I highly recommend using the <a target="_blank" rel="ugc" href="https://github.com/Bing-su/adetailer">ADetailer</a> extension for better faces and eyes.For v2, I merged v1 with some NoobAI models, <a target="_blank" rel="noopener noreferrer nofollow" href="https://civitai.com/user/GarejiSama">@GarejiSama</a>'s <a target="_blank" rel="ugc" href="https://civitai.com/models/1649312?modelVersionId=1866796">AdAstra</a> and <a target="_blank" rel="noopener noreferrer nofollow" href="https://civitai.com/user/Manityro">@Manityro</a>'s <a target="_blank" rel="ugc" href="https://civitai.com/models/1732765/vermilion">Vermilion</a>, making v2 much more compatible with NoobAI based LoRAs. The baked in VAE is Anzhc's, which you can get <a target="_blank" rel="ugc" href="https://huggingface.co/Anzhc/Anzhcs-VAEs/blob/main/SDXL%20Anime%20VAE%20Dec-only%20B3.safetensors">here</a>, gives a much sharper look to images, reduced noise, and makes colors less bland overall. LustrousMix v2 generations are more colorful than v1's, with more contrast and sharpness. There is a slight art style change as well, with more mature faces.

Files with SHA256: 5c0e3ed489766001c68f1fb983375179a4c9dc614bfe1cf29db59f7fa8268f1e