PixAI Membership
PixAI
PixAI
Generate
Sign in
Home
Generation
Models
Studio
NEW
Contest
Create with Mio.2
AI Tools
NEW
Guide

User LoRA

SD

Potion / ぽ~しょん Art Style 0.1

#Original#Mature Female#Girl#Potion#Aina Winchester#Shinohara Mari#Akugami Mai#Minamoto no Raikou
AI Tools
  • AI Anime Generator
  • AI Image Generator
  • AI Animation Generator
  • Anime Figure Generator
  • Character Sheet Generator
  • Magazine Cover Generator
  • Plushie Generator
  • Giant Statue Generator

Discover

  • Trending Tags
  • Ranking
  • Model Market
  • Contest
  • News

About

  • Guide
  • PixAI Blog
  • Tsubaki.2
  • Meet Mio, Our Mascot
  • Content Rules

Pricing & Help

  • Membership
  • Buy Credits
  • Contact

Mobile App

  • Get the PixAI app
  • App Store
  • Google Play
© 2026 PixAI
  • Terms of Service
  • Privacy Policy
  • Copyright Policy
  • Attributions
  • Specified Commercial Transactions Law
Potion / ぽ~しょん Art Style 0.1 - AI Model cover image
Trained by
h3

hs3

Train Your Own LoRA
Uploaded at
Sep 14, 2023, 10:01 PM
Uses
9.8k
Earned
150,410
Reviews
No reviews
Trigger Words

Potion_ぽ~しょん_art_style

Permissions
  • Allow image generation & sharing
  • Allow users to download this model
  • Allow commercial use of generated images

Recommended LoRAs

Yuzuna Hiyo (柚子奈 ひよ) - Style - AI Model cover image

User LoRA

Yuzuna Hiyo (柚子奈 ひよ) - Style
9
Yuzuna Hiyo style LORA - AI Model cover image

User LoRA

Yuzuna Hiyo style LORA
22
Z Mogudan Stile - AI Model cover image

User LoRA

Z Mogudan Stile
18
dtdfhww - AI Model cover image

User LoRA

dtdfhww
0

Description

Original Style: potion / moudamepo Nickname: ぽ~しょん Former nickname: ポーション Pixiv: https://www.pixiv.net/en/users/565527 Danbooru: https://danbooru.donmai.us/posts?tags=potion_%28moudamepo%29

Explanation: This Lora has been trained with 100 drawings by the artist "Potion" that are as diverse as possible and include everything from old sketches to more recent game CGs of Valkyria, in order to resemble Potion's style as closely as possible.

Note: This Lora works especially well with reference images of whole-body drawings and higher generation sizes.

Recommended model: After some tests, the "AbyssOrangeMix2 - Hardcore" model seems to work by far the best, almost at the level of an original game CG from a company, although it takes a long time to generate it. The "Hassaku" model is much faster and you can also work well with it, but the results are not as good though.

Recommended Prompts: In general, most prompts related to outfits and clothing seem to work best, also shorter prompts have worked much better for me so far.

Recommended settings: I tended to use default settings and didn't do much testing, so others may work better. Sampling Steps: 23 Sampling Method: Euler CFG Scale: 6 Size: 512x768

Comments