Model Information

Latex Diffusion V1 (reupload) - Image 1
Latex Diffusion V1 (reupload) - Image 2
Latex Diffusion V1 (reupload) - Image 3
Latex Diffusion V1 (reupload) - Image 4
Latex Diffusion V1 (reupload) - Image 5

Description

New model: Abyss+Bondage+Latex (safetensor)
Abyssorangemix2 + Bondage V10 (on 0.15)
Result + Latex (on 0.3)
(needs the "vae-ft-mse-840000-ema-pruned" to correct the greyish color)

So, after training my latex model, I was disappointed that it was so unstable and I thought it was a failure.
But then I merged it with Berrymix and I was amazed by the results. It was as if merging it had fixed all problems with it.
Then I also made some merging experiments and made the ArtEros+Berrymix+Latex model which I am using most of the time.
In both cases I had merged the raw latex model at the end 50/50 with the other model. So if you want to make your own, this might be a good starting point.

The training data was rather limited with 42 images (but a new version with a bigger dataset might be coming...).
This resulted in:
- basically 3 base outfits
- a standard latex catsuit
- a standard latex catsuit with a corset on top of it
- only latex corset with latex stockings on straps connected with the corset and long sleeved gloves; breasts are exposed
- breasts are usually big and round (due to training data)
- once merged, the resulting models have a heavy bias towards shiny outfits and slim waists (even without corset), even without any of the trigger words

Latex Diffusion V1 (reupload)

1.2Kdownloads
Download Model

Model Details

Type
AI Model
Task
text-to-image
Subtype
Safetensors / Checkpoint AI Model
Created
Updated
June 3, 2026

Tags

Related Models