Mydei — Character LoRA (Animagine XL / SDXL)

High-quality character LoRA for Mydei, trained on a clean dataset with high-resolution images. Designed specifically for Animagine XL 3.1 / SDXL anime checkpoints.

It captures his sharp facial features, amber-red eyes, signature hair gradient, red torso tattoos, and armor with high consistency, while maintaining excellent outfit flexibility.

⚙️ Recommended Settings

ParameterRecommended ValueBase ModelAnimagine XL 3.1 / Any SDXL Anime ModelLoRA Weight0.8 (optimal range: 0.7 – 0.9)Trigger WordmydeiHelper Tags1boy, solo, amber eyes, blond hair, red markings on torsoCFG Scale5.0 – 7.0Sampling Steps25 – 30SamplerEuler a or DPM++ 2M KarrasResolution832x1216 (Portrait) / 1216x832 (Landscape) / 1024x1024

Key Features

• Outfit & Style Flexibility: Trained with clean tag separation. Easily swap his canon armor for leather jackets, beachwear, modern clothes, or casual outfits without bleeding artifacts.

• Pose Stability: Excellent performance on full-body shots, action poses, and dynamic angles.

• Facial Consistency: Accurately retains eye color, expression, and hair shape across various weights.

Example Prompts

Positive Prompt Template:

Plaintext

score_9, score_8_up, score_7_up, <lora:Mydei-000008:0.8>, 1boy, mydei, solo, looking at viewer, upper body, gentle smile, unbuttoned shirt, casual clothes, outdoor, masterpiece, best quality, highly detailed

Negative Prompt Template:

Plaintext

score_4, score_3, score_2, score_1, lowres, bad anatomy, bad hands, missing fingers, extra digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, blurry

If you enjoy this model, please consider leaving a like, review, or sharing your generations below! Feedback is always appreciated.