Look, we all know what this is by now.

My LoRAs are trained on large datasets with a focus on specific characters appearing together.

You can find the exact prompts used in the sample metadata.

A quick note on how these work:

• Multi-Character: They are trained to handle characters together. I DO NOT use regional prompting or similar for any of my samples. This will result in outfit bleed a portion of the time. Particularly when the characters are same sex and share certain tags.

• Individual Use: They are ALSO trained on the individual characters and function perfectly fine as such (maybe even slightly more accurately). Again, samples are attached.

• Regional Prompting: While I don't use it, these are compatible with regional prompting extensions if you want to ensure clean outfits every time.

• I will be adding more character combos as different versions here, and other shows as well.

Throw me a like. Or don't. Whatever. I'll probably get bored making these in a few weeks anyway.