Ahli PixAI
PixAI
PixAI
Generate
Sign in
Laman Utama
Jana
Model
Studio
NEW
Kontes
Cipta dengan Mio.2
Kotak Alat
NEW
Guide

User LoRA

XL
Gaya

ACE GLITCH | MIKUNEKI Style Lora v1.2

Cipta

  • Penjana Imej AI
  • Penjana Animasi AI
  • Kotak Alat
  • Penjana Tematik
  • Latih LoRA
  • Ejen Mio.2
  • Studio

Temui

  • Tag Popular
  • Ranking
  • Pasaran Model
  • Kontes
  • Berita

Tentang

  • Guide
  • Cara Menggunakan PixAI
  • Tsubaki.2
  • Kenali Mio
  • Peraturan Kandungan

Harga & Bantuan

  • Keanggotaan
  • Pek Kredit
  • Kontak

Aplikasi Mudah Alih

  • App Store
  • Google Play
© 2026 PixAI
  • Syarat Perkhidmatan
  • Dasar Privasi
  • Dasar Hak Cipta
  • Undang-undang Transaksi Komersial Tertentu
ACE GLITCH | MIKUNEKI Style Lora v1.2 - AI Model cover image
Dimuat naik pada
13 Nov 2025, 10:22 PG
Kegunaan
7.9k
Ulasan
Cemerlang  (2)
Kebenaran
  • Allow image generation & sharing
  • Benarkan pengguna memuat turun model anda
  • Kegunaan komersial

Recommended LoRAs

Japanese DollLikeness LoRA - AI Model cover image

User LoRA

Japanese DollLikeness LoRA
894
Sorey Tales of Zestiria the X (Pony XL) - AI Model cover image

User LoRA

Sorey Tales of Zestiria the X (Pony XL)
9
Cute little fang XL - AI Model cover image

User LoRA

Cute little fang XL
9
 ye shunguang void hunther Ex - AI Model cover image

User LoRA

ye shunguang void hunther Ex
1

Keterangan

Ace glitch/Mikuneki style lora - v1.2(Probably the best one there is, i was honestly baffled by how well it generalizes with other characters)Looking for a great LoRA that will apply your favorite artists' style? look no further! for some god forsaken reason the unhinged settings i used for it work, and they worked well!So, as stated above. it's basically a LoRA for Ace glitch/Mikuneki's style. it's probably the only LoRA of mine that i actually use a lot, it can do pretty much any character you feed it and it doesn't clash with other LoRAs at all. managed to use it with 3 other stacked loras with absolutely 0 quality drop.Recommended weight is 1.8, i don't know why it's so timid but i guess that's the tradeoff for perfect generalization. and i guess it gives you more leverage for controlling it. the keywords i used in training were "1girl, masterpiece, best quality, sharp lines" so you might REALLY want to use them.v1.2 is most likely final as i don't see what else i could really do with it, im actually really happy with the results which is pretty rare.Feel free to request any other LoRAs whether concept or style!(and share your images, pweaase~ :3)Oh also, important note. I removed Ace Glitch's watermark for a few reasons, the most important being:I'm afraid some AI slop patreons might try to impersonate miku/ace, whatever you get the gist. don't want to taint their name and style.Watermarks suck for training. models sometimes go batshit because of them since they're hard to classify.To avoid ace's and miku's wrath. Although if either of the two are reading this, you are completely free to DM me and i'll take down the model. let's be respectful here.

Komen2