---
license: mit
pipeline_tag: text-to-image
tags:
  - comfyui
  - krea2
  - moss-image
  - anima
  - illustrious
  - cinematic
  - moody
---

# Animosity 🖤

My dark, moody, cinematic look, on four base models. Low-key light, deep color, crisp detail. Built for ComfyUI.

- **MOSS Image:** AniMOSSity 🆕
- **Krea 2 Turbo**
- **Anima**
- **Illustrious**

---

## ✋ Which file do I grab?

**MOSS Image: AniMOSSity** 🆕 (`moss/`)

| File | Size | Pick it if |
|---|---|---|
| **AniMOSSity_int8_convrot.safetensors** | 5.4 GB | 💡 most people. Runs on RTX 30 series and newer |
| AniMOSSity_bf16.safetensors | 10.2 GB | you want the reference file |
| Animosity_MOSS LoRA.safetensors | 129 MB | you want the look on your own MOSS SFT setup or mixed with other LoRAs (goes in `loras/`) |

- AniMOSSity is the LoRA baked into MOSS Image SFT at 0.9. The int8 renders the same picture as the bf16.

**Krea 2 Turbo** (`krea2/`)

| File | Size | Pick it if |
|---|---|---|
| **Animosity_Krea2_V1.0_int8_convrot.safetensors** | 14.1 GB | 💡 most people. bf16 quality at half the size |
| Animosity_Krea2_V1.0.safetensors | 26.3 GB | you have 32 GB of VRAM and want the reference |
| Krea-Animosity_V1.0-fp8.safetensors | 14.1 GB | your setup runs fp8 but not int8 |
| Animosity_krea2_V1.0-q8_0.gguf | 15.6 GB | you load models through GGUF |

**Anima** (`anima/`)
- animosity_animaV20Base.safetensors (4.2 GB)

**Illustrious** (`illustrious/`)
- Animosity_Illustrious_v1.1.safetensors (7.3 GB). A full checkpoint: it goes in `checkpoints/`.

📂 Where they go:

```
ComfyUI/
└── models/
    ├── diffusion_models/
    │   └── (MOSS, Krea 2 and Anima files)
    ├── checkpoints/
    │   └── (Illustrious)
    └── loras/
        └── (Animosity_MOSS LoRA)
```

💡 My **LC Model Optimizer** (LC123 pack) picks the right file for your card and downloads it for you.

---

## Settings

**MOSS Image**
- Steps: 25
- cfg: 2.5
- Sampler: er_sde
- Scheduler: simple
- Skip Layer Guidance on layer 22, 0 to 0.5
- Text encoder: qwen3.5_9b_int8_convrot. VAE: flux2-vae
- ⚠️ MOSS needs a ComfyUI with native MOSS Image (newer than 0.41.3).

**Krea 2 Turbo**
- Steps: 10
- cfg: 1.0
- Sampler: Euler
- Scheduler: Simple
- VAE: qwen_image_vae

**Anima**
- Steps: 40
- cfg: 5
- Text encoder: qwen_3_06b_base. VAE: qwen_image_vae
- **Seeds3 & Bong_tangent:** winner
- **Euler & Simple:** pulls out a lot of detail
- **Seeds3 & Simple:** pulls off a ZIT look quite well 🤔

**Illustrious**
- **Dpmpp++ Karras:** pulls out the color
- **Euler & Simple:** pulls out a lot of detail
- **Seeds3 & Simple:** pulls off a ZIT look quite well 🤔
- **Deis & Simple:** thumbs up 👍
- **Deis & sgm_uniform:** oh wow 😍

---

Find it on CivitAI [Here](https://civitai.com/models/2596298)

Instagram: [synth.studio.models](https://www.instagram.com/synth.studio.models/)

Buy me a ☕ [HERE](https://ko-fi.com/lonecatone)

This represents many hours of work. If you enjoy it, please 👍 like, 💬 comment, and feel free to ⚡ tip 😉

"True Nothing is. Permitted Everything is"- Yoda Auditore, Assassin's Wars