This post now contains 3 different Anima LoRAs, each converted from styles we originally used in SD / NovelAI.They are grouped together because they come from the same circle and workflow — but each version has a different vibe.All versions are trained specifically for Anima (DiT). ----------------------------------------------------------------------<h3 id="4x2-hanpi">✦ 4X2 – Hanpi </h3>Inspired by Hanpi. Onineko based style with fangdongye, karo, and kedama milk. This version pushes softness and cute energy. It is dreamy, moe and eye-heavy.Colors are lighter overall, with a softer, slightly washed-white rendering in both shading and highlights. The lighting feels airy and glowing. Eyes are larger and more luminous, gradients are smoother, and often with layered gradients and strong shine. Linework can include subtle colored outlines instead of pure black, which adds to the dreamy look. ----------------------------------------------------------------------<h3 id="4x1-ayori-yu-based">✦ 4X1 – Ayori / Yu-based </h3>This version feels more painterly, with slightly softer and sometimes blurred brush edges instead of hard linework.Lighting is much stronger here — it naturally produces bold light and shadow contrast.Compositions also tend to feel more dynamic and cinematic. It works especially well for artistic anime illustrations rather than flat character sheets.White stockings / white fabric textures render particularly nicely in this version — the material shading and light interaction feel more dimensional. ----------------------------------------------------------------------<h3 id="4x0-aaac-kupa">✦ 4X0 – AAAC Kupa</h3>This version leans toward solid, confident linework and a slightly dark gothic lighting mood.The lines feel firm and structured rather than soft or painterly. Eyes are typically long and sharp, giving characters a more intense presence.Colors sit on the darker side of the spectrum — not overly bright or washed out. They’re moderately saturated but with lower overall luminance, so the palette feels rich and moody instead of glowing.

Style - 4x0 Aaalice family Style [Anima]

anima_4x1style_epoch80.safetensors

4x1 Ayori Yu V1.0

4x2 Hanpi v2.0

4x1 Ayori Yu V2.0

4x2 Hanpi V1.0

4x1Style

4x0 AAAC Kupa V1.0

this lycoris is not mine, here the desc :This is an experimental LoRA version trained on Anima.Anima is a new DiT model focused on anime concepts. You can find it on Hugging Face at <a target="_blank" rel="ugc" href="https://huggingface.co/circlestone-labs/Anima">circlestone-labs/Anima · Hugging Face</a>The dataset consists of images that I recently generated using Noob. Trigger word@4x0styleRecommended strength is 1.0The Anima model structure is completely different from SDXL. Please only use the Anima version for Anima model generations. Use the NoobAI version for SDXL models, like Noob and IL.

[Anima] Style - 4x0 Style

This is an experimental LoRA version trained on Anima.

Anima is a new DiT model focused on anime concepts. You can find it on Hugging Face at circlestone-labs/Anima · Hugging Face

The dataset consists of images that I recently generated using Noob.

Trigger word

@4x0style

Recommended strength is 1.0

The Anima model structure is completely different from SDXL.
Please only use the Anima version for Anima model generations.
Use the NoobAI version for SDXL models, like Noob and IL.

Files with SHA256: 8aec19c81dbbfb78f0053c704f2f6144cde9d49077180e656ada46a74cedeabd