Ada Wong | Pony Diffusion XL
Bring the mysterious and iconic Ada Wong from Resident Evil into your Pony Diffusion XL generations!
This character LoRA was trained using 120 carefully selected and manually captioned images, featuring a variety of outfits, expressions, poses, and environments.
Designed to capture Ada's recognizable appearance while allowing flexibility across different scenarios, from her signature red dress and tactical outfits to casual fashion, elegant evening wear, swimwear, lingerie, and mature-themed artwork.
Model Information
Base Model: Pony Diffusion XL
Model Type: Character LoRA
Training Platform: CivitAI
Training Dataset: 120 images
Training: 10 epochs / 2,000 total steps
Trigger Word:
ada_wongCaptioning: Manually written Danbooru-style tags
Recommended Usage
Trigger word: ada_wong
Recommended LoRA strength: 0.55–0.60
During testing, lower LoRA strengths produced better overall results while maintaining Ada's recognizable appearance. Higher strengths may introduce unwanted anatomy or composition issues.
Suggested Pony quality tags:
score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up
Example Prompt
score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, ada_wong, 1girl, solo, short black bob hair, brown eyes, red halter dress, black thighhighs, high heels, full body, standing on a rainy city street at night, confident expression, cinematic lighting, detailed background
Suggested LoRA strength: 0.6
Feel free to adjust the strength and generation settings to suit your preferred Pony checkpoint.
What Can This LoRA Do?
This model was tested across a variety of outfits and scenarios, including:
Ada's iconic red dress and classic Resident Evil-inspired looks
Tactical and spy-inspired outfits
Casual clothing and alternative fashion
Elegant evening gowns and formal wear
Swimwear and lingerie
Mature-themed artwork
Different poses, expressions, camera angles, and environments
Results depend on your checkpoint, prompt, LoRA strength, and generation settings.
Showcase Images & ComfyUI Workflows
Showcase images include embedded generation metadata!
You can inspect the prompts and settings used to generate the examples.
Original PNGs retaining their ComfyUI workflow metadata can also be downloaded and dragged directly into ComfyUI to load the generation workflow.
Required checkpoints, LoRAs, and custom nodes may need to be installed separately.
Other Versions
This is the Pony Diffusion XL edition of my Ada Wong character LoRA.
Separate versions have also been created for:
Illustrious XL
Anima
Each version has been trained and tested separately for its respective base model.
About This Model
Created by Xander's Workshop.
If you enjoy the model, consider leaving a like, sharing your creations, or following Xander's Workshop for more character LoRAs!
Have fun creating! ❤️
Description
Version 1.0 | Pony Diffusion XL
Initial release of Ada Wong for Pony Diffusion XL.
Trained on 120 manually captioned images across 10 epochs (2,000 training steps), featuring a variety of outfits, poses, expressions, and environments.
Tested across classic Resident Evil-inspired outfits, casual clothing, formal wear, swimwear, lingerie, and mature-themed scenes.
Recommended LoRA strength: 0.55–0.60
Testing showed that lower strengths provided better overall image quality and anatomy while maintaining Ada's recognizable appearance.
Trigger word: ada_wong
Supports both SFW and NSFW generations. See the main description and showcase images for example prompts, recommended settings, and embedded ComfyUI workflow metadata.












