Sign In

Leon - Callisto Protocol

81

Updated: Oct 26, 2026

character

Not recently loaded

Download

1 variant available

SafeTensor

435.34 MB

Verified:

Type
LoRA
Stats

59

24

4

Reviews
Published

Jun 23, 2025

Base Model
Usage Tips
Clip Skip: 1
Hash
AutoV2FDF4904D4B
Trigger Words
leon
00030-3802501054.png

This LoRA is designed to generate images of the Protocol Callisto character Leon Ferris.

Available for Krea2, Anima and Illustrious.

Recommended LoRA Strength Krea2: [0.7 – 1.0]
Recommended LoRA Strength Anima: [0.8 – 1.0]

Trigger words: leonn

Base Outfit:


KREA2: leonn, A male with a buzz cut. He wears a navy tactical jumpsuit with padded sleeves and a high ribbed collar, dark olive tactical armor with a heavy segmented chest plate and black harness straps over the shoulders, gold and silver armbands on the upper arms, an armored gauntlet with red lights on one forearm, black tactical gloves, a black utility belt with pouches and a triangular buckle, a large utility pouch strapped to the thigh with a leg strap, olive and brown knee pads and black combat boots with silver buckled straps,

ANIMA: leonn, 1boy, solo, male focus, mature male, buzz cut, very short hair, navy jumpsuit, padded sleeves, high collar, olive armor, chest plate, black harness straps, gold armband, armored gauntlet, black gloves, utility belt, belt pouch, triangular belt buckle, thigh pouch, leg strap, knee pads, black boots,

Piece by piece: KREA2

  • Collar and shoulders: a high ribbed collar, navy padded sleeves and black harness straps over the shoulders

  • Chest: a heavy dark olive chest plate with segmented panels and bolts

  • Arms and hands: gold and silver armbands on the upper arms, an armored gauntlet with red lights on one forearm, black tactical gloves

  • Belt and legs: a black utility belt with pouches and a triangular buckle, navy trousers with dark side panels, a large utility pouch strapped to the thigh with a leg strap, olive and brown knee pads

  • Boots: black combat boots with silver buckled straps and thick treaded soles

  • Back: a heavy olive and brown back plate with ribbed panels, a central spine ridge and vents

Piece by piece: ANIMA

  • Collar and shoulders: high collar, padded sleeves, black harness straps,

  • Chest: olive armor, chest plate, segmented panels,

  • Arms and hands: gold armband, silver armband, armored gauntlet, black gloves,

  • Belt and legs: utility belt, triangular belt buckle, belt pouch, thigh pouch, leg strap, knee pads,

  • Boots: black boots, buckled straps,

  • Back: from behind, facing away, back plate, vents,

Tips:


KREA2: write your prompt as natural sentences, starting with “leonn, A male with a buzz cut.” Then add the style (“3D video game render” for the game look, “digital illustration” for an illustration), the scene, and the outfit sentence above. Quality tags and negative prompts are not needed. If the armor shows up when you don’t want it, describe the new clothes explicitly (for example “a black leather jacket, no armor”).

ANIMA: use Danbooru-style tags with the trigger word first, then one short natural-language sentence (the LoRA was trained with mixed captions). Quality tags: “masterpiece, best quality”. Negative prompt: “worst quality, low quality, score_1, score_2, score_3, blurry, jpeg artifacts, bad anatomy, bad hands, extra fingers, missing fingers, deformed, watermark, signature, text, username”. The dataset is 3D game renders tagged “3d, realistic, video game”: use those tags for the game look. For anime or illustration styles, leave them out, add “anime coloring, cel shading, clean lineart” and add “3d, realistic, photorealistic, photo” to the negative.

BOTH: if the 3D game look overpowers an illustration prompt, lower the strength to about 0.8. If the face loses likeness in full-body shots, raise the strength or use hires fix or a face detailer.

Notes: 31 training images, mostly close-ups and upper body, with fewer full-body, side and back views. There are no shirtless images in the dataset, so body type comes from the base model. Many images are dark or have a plain white background.

Please check the gallery images to see the exact prompts and settings I used to get an idea of how it behaves. I’d love to see your generations, so feel free to share your results in the comments!

This LoRA was trained using the following base models:

  • krea2_raw_bf16

  • anima-base-v1.0

Tested in Forge Neo UI.