Sign In

Loki - [Danmachi]

0

Updated: Sep 26, 2026

characterred hairdanmachiloki

Not loaded

Download

1 variant available

bf16 SafeTensor

loki_danmachi_kurbaga_v1.safetensors

BF16, good balance • 66.23 MB

Verified:

Type
LoRA
Stats

62

0

Generation

Not loaded

Reviews
Published

Sep 26, 2026

Base Model

Anima

Training
Steps: 814
Epochs: 74
Usage Tips
Strength: 1
Hash
AutoV2
3634F02212
Trigger Words
loki_kurba
loki_kurba, red hair, medium hair, ponytail, hair ornament, closed eyes, flat chest,
blue bandeau, black sleeves, navel, black shorts, belt, black thighhighs, blue shoes, high heels,
default creator card background decoration
Uploads - 20

20

Likes - 577

577

Downloads - 4318

4.3K

License:

Anima

The Anima Model is licensed by CircleStone Labs LLC. Copyright CircleStone Labs LLC. IN NO EVENT SHALL CIRCLESTONE LABS LLC BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.

Built on NVIDIA Cosmos

ComfyUI_00236_.png

This is my first locally trained Anima-based LoRA!

Appearance & Outfit tags:

loki_kurba, red hair, medium hair, ponytail, hair ornament, closed eyes, flat chest,
blue bandeau, black sleeves, navel, black shorts, belt, black thighhighs, blue shoes, high heels,

If you want the eyes to be open, replace closed eyes with red eyes.

I've had mixed results while using quality tags. Here are a few that I've tested and found to work across multiple checkpoints.

Generally okay to use with any checkpoint:

masterpiece, best quality, score_9, score_8, score_7, absurdres, highres, newest, 

Best with higher-detail checkpoints:

hyper realism, photorealistic,

The realism tags improved overall quality while preserving the anime style.

The lighting was the most notable difference when using these tags.

Recommended settings:

  • Sampling Steps: 15-35

  • CFG Scale: 3.5-4

Check out the suggested resources to get similar results to the preview images!

This LoRA was trained on a small, mixed-quality dataset.

NSFW generations performed quite well with this LoRA.

I would love to see what you create with this LoRA. If you have any feedback, I'd love to hear it.