---
license: openrail++
tags:
  - text-to-image
  - stable-diffusion
  - stable-diffusion-xl
---

<style>
    .title {
        font-size: 48px;
        text-align: center;
        letter-spacing: 0.1em;
        padding: 20px 0px;
        background: transparent;
    }

    .title span {
        background: -webkit-linear-gradient(66deg, #00a6de, #9b70ff);
        -webkit-background-clip: text;
        -webkit-text-fill-color: transparent;
    }

    .sample-image{
        width: 100%;
        height: auto;
        border-radius: 8px;
        transition: transform .2s; 
    }

    .sample-image:hover{
        transform: scale(1.03);
    }
</style>

<h1 class="title"><span>SakumomoXL</span></h1>

<hr>

<table>
    <tr>
        <td>
            <img class="sample-image" src="https://huggingface.co/KidiXDev/sakumomoXL/resolve/main/assets/img04.png" alt="img4">
        </td>
        <td>
            <img class="sample-image" src="https://huggingface.co/KidiXDev/sakumomoXL/resolve/main/assets/img02.png" alt="img2">
        </td>
        <td>
            <img class="sample-image" src="https://huggingface.co/KidiXDev/sakumomoXL/resolve/main/assets/img01.png" alt="img1">
        </td>
        <td>
            <img class="sample-image" src="https://huggingface.co/KidiXDev/sakumomoXL/resolve/main/assets/img03.png" alt="img3">
        </td>
    </tr>
</table>

<hr>

This model was created using a simple merge of several other checkpoints. you can find the base model [here](https://civitai.com/models/316882?modelVersionId=382513).

## What is this?

This model aims to offer more flexibility and stability in the art style while keeping it consistent with the original design. It balances creative freedom with a steady visual framework to preserve the style and allow for smooth transitions and refined adjustments.

## Merge configuration

v1-
Base: [MomoiroPony_v1.4](https://civitai.com/models/316882?modelVersionId=382513)

- [HassakuXL-H_v1.3](https://civitai.com/models/376031/hassaku-xl-hentai?modelVersionId=559994)
- Ratio: 0.76
- CLIPMerge Ratio: 0.83
- BakedVAE: [xlVAEC_f2](https://civitai.com/models/152040?modelVersionId=719135)
- Output: sakumomoXL_v1

v2-
Base: sakumomoXL_v1

- [Sweet-mix_XLPony_v1.2](https://civitai.com/models/18927?modelVersionId=771458)
- Ratio: 0.43
- CLIPMerge Ratio: 0.23
- BakedVAE: ?
- Output: sakumomoXL_v2

Tools used for merging: [ComfyUI](https://github.com/comfyanonymous/ComfyUI)

## Usage examples:

Quality Prompt - Make sure you include this:

```
score_9, score_8_up, score_7_up
```

Positive Prompt:

```
score_9, score_8_up, score_7_up, source_anime, 1girl, solo, cute, light smile, red hair, blue eyes, outdoors, face focus, rating safe
```

Negative Prompt:

```
(score_4,score_3,score_2,score_1:1.3),  bad anatomy, deformed anatomy, realistic, source_furry, watermark, animal ears, heart-shaped pupils, abs, desaturated, noise, blur
```

Recomended Image Size:

```
832x1216 (Portrait)
768x1344 (Portrait)
```

CLIP Skip:

```
2
```

Sampler:

```
Euler a
```