trixxi

Trixxi — triXX.i Character Model

Trixxi is a stylized character LoRA/model of triXX.i — red bob, green square frames, glitter blue eyeshadow, double stacks, VVS diamond grill, LUVI-era neon aesthetic. Trained for consistent character generation, tattoo flash, and poster/ad creative.

Model Details

• Model Name: Trixxi
• Version: 1.0
• Creator: Alana Battaile / triXX.i
• Model Type: Character LoRA for SDXL (compatible with SDXL 1.0, Pony, Juggernaut)
• Trigger Word: triXX.i, Trixxi
• Base Model: stabilityai/stable-diffusion-xl-base-1.0
• Training Resolution: 1024x1024
• License: CreativeML Open RAIL++-M

Intended Use

Intended for:

• Consistent character renders of Trixxi
• Tattoo flash art, streetwear ads, LUVI brand creative
• Stylized portraits, posters, Instagram story ads

Trigger + style prompt: triXX.i, red bob haircut, green square glasses, glitter blue eyeshadow, vvs diamond teeth grill, double nose piercings

Add-ons:

• tattoo flash sheet, bold traditional linework, off-white paper
• neon city billboard at night, LUVI ad, cinematic lighting
• fashion editorial, close-up portrait, sharp focus

Not intended for:

• Real person impersonation outside of the fictional Trixxi character
• NSFW / explicit content

How to Use

Diffusers

from diffusers import DiffusionPipeline
import torch

pipe = DiffusionPipeline.from_pretrained(
    "stabilityai/stable-diffusion-xl-base-1.0",
    torch_dtype=torch.float16
).to("cuda")

pipe.load_lora_weights("AlanaSky/trixxi", weight_name="trixxi_v1.safetensors")
# adjust weight: 0.8 - 1.0 recommended
pipe.fuse_lora(lora_scale=0.9)

prompt = "triXX.i, red bob, green square glasses, glitter blue eyeshadow, vvs diamonds covering top teeth, LUVI neon jacket, portrait, high detail"
negative_prompt = "blurry, low quality, distorted teeth, extra fingers"

image = pipe(prompt=prompt, negative_prompt=negative_prompt, num_inference_steps=30, guidance_scale=7.5).images[0]
image.save("trixxi.png")

A1111 / ComfyUI / Forge

1. Download trixxi_v1.safetensors
2. Place in /models/Loras/
3. Use: <lora:trixxi_v1:0.9> triXX.i, [your prompt]

Recommended settings:

• Weight: 0.85 – 1.0
• CFG: 6.5 – 8
• Steps: 25–35
• Sampler: DPM++ 2M Karras / Euler a
• Hires fix: 1.5x with 0.4 denoise

Training Data

• Dataset: [e.g., 45 curated images of Trixxi — portraits, flash sheets, neon ad shots]
• Captioning: Manual + WD14 + BLIP, with trigger triXX.i
• Steps: [e.g., 2500 steps, 3 epochs]
• Optimizer: AdamW8bit, lr 1e-4
• Training Tool: Kohya_ss / OneTrainer

Limitations & Bias

• Teeth/diamond grill can warp at low weight or low steps — increase weight to 0.9+ and add sharp teeth, vvs diamonds to the prompt.
• Green glasses color can bleed if you add other green clothing without weighting: (green square glasses:1.2)
• Model is overfit to red bob — use blonde hair with a red hair negative if you want variants.

Samples

License

This model is licensed under CreativeML Open RAIL++-M. Commercial use is allowed for the model weights, but you are responsible for generated content. The Trixxi character and LUVI branding retain their original trademark.

Credits

• Character: triXX.i / Trixxi
• Model by: Alana Battaile
• Contact: [your Instagram handle / email]

If you like the model, please like and share renders!

@misc{alana_battaile_2026, author = { Alana Battaile }, title = { trixxi (Revision d16253c) }, year = 2026, url = { https://huggingface.co/AlanaSky/trixxi }, doi = { 10.57967/hf/10083 }, publisher = { Hugging Face } }