— FOXTONTONTON · MUSCLE × LATEX × GROWTH —
You can find me on any of these platforms — same name everywhere: FOXTONTONTON. Follow wherever you already hang out, and you won't miss the next chapter, illustration, or animation.
🎬 X/Twitter | 🖼️ DeviantArt | ☕ Ko-fi | 💎 Civitai | 🦋 Bluesky
Everything free is downscaled and censored — the full HD, uncensored downloads (images & videos) live on my Patreon. If you'd like to support my work, that's the best place to do it. You can also buy me a coffee on Ko-fi. Thank you so much!
About this version
SD 1.5 Edition — Official Base LoRA (Version Changes & Notes)
Version Overview & Highlights
Trained Natively on Official SD 1.5 (
v1-5-pruned-emaonly): Built from scratch onstable-diffusion-v1-5/stable-diffusion-v1-5(v1-5-pruned-emaonly.safetensors) rather than a stylized merge. Training on the clean canonical SD 1.5 base ensures broad compatibility and predictable weight blending across photorealistic, semi-realistic, and stylized SD 1.5 checkpoints without introducing base-model bias.Trigger Word:
blacklatexmaleRecommended Weight:
0.65 – 0.85(best balance around0.75on the final Epoch 12 checkpoint; intermediate checkpoints saved every 2 epochs work well if you want lighter latex coverage or multi-LoRA stacking).Recommended Inference Settings:
Prompt Prefix:
blacklatexmale, 1boy, solo, hyper muscular male, black latex catsuit, skintight, glossy latex, dramatic rim light(Note: Unlike Pony models, do not use
score_9, score_8_uptags with this SD 1.5 version.)CLIP Skip:
1for photorealistic SD 1.5 checkpoints (or2when using stylized/anime SD 1.5 merges)Sampler / Scheduler:
DPM++ 2M SDE+KarrasorEuler a,25–32steps, CFG5.0 – 7.0Resolution & Upscaling: Generate at
512×768,768×512, or512×512, and pair withHires.fix(1.5× – 2.0×) +ADetailer/FaceDetailerfor crisp facial micro-details on full-body shots.
Author's Note (Final SD 1.5 Version)
This is expected to be the final version for the SD 1.5 model branch. Overall results turned out much better than expected—latex specular reflections, muscular proportions, and pose consistency hold up remarkably well across SD 1.5 checkpoints. However, because of SD 1.5's lower native resolution, I still strongly recommend adding face and penis detailing passes (ADetailer / FaceDetailer) (together with Hires.fix) for crisp facial features and anatomical detail.
Description
This is the first experimental release of BlackLatexMale.
The model was trained using Kohya_ss on Google Colab as a learning and testing project. The current version focuses on muscular male characters wearing black shiny latex and rubber bodysuits.
While this version already produces solid results, it is still an early test build and may have inconsistencies in anatomy, latex reflections, or pose flexibility in some generations.
Future updates may include:
improved anatomy consistency
better latex material rendering
stronger lighting and reflection details
expanded pose and body variation
improved NSFW support and compatibility
Thank you for trying this LoRA and supporting future development.








