Create

Generate Similar

Generation Mode Choose how the image should start.

Prompt

Master Prompt:
Hyper-detailed Japanese manga editorial poster, extreme close-up 
face portrait, 9:16 phone wallpaper format. Character shows 
battle damage — cuts, blood, sweat on face. Expression ranges 
from rage to cold menace to sardonic dominance. Dynamic spiky 
or wild hair. Supernatural eye glow (red / purple / blue — 
character dependent).

CORE AESTHETIC: Weekly Shonen Jump cover meets traditional 
Japanese woodblock print. NOT Western comic — authentic 
Japanese manga inking, precise halftone dots, sumi-e brush 
calligraphy, washi paper collage layers.

BASE TEXTURE: Aged washi paper with foxing marks + manga 
halftone grain overlay at 30% opacity + fine noise texture.

BACKGROUND SYSTEM (layered, all 4 rules apply):
- Torn paper collage strips at angles — 3 to 5 layers deep
- Diagonal ink slash/scratch lines radiating from character
- Manga panel fragment inserts in corners (monochrome line art)
- Character-specific symbol faintly at 20% opacity behind figure
  (spiral / torii / chain / tomoe — match character)
- Ink splatter + speed lines (効果線) near action points
- Deep vignette — near black at all four corners

COLOR RULES:
- ONE dominant color per poster (red / dark red / deep purple / 
  ink black) — takes 60% of palette
- Bone white (#F0EBE0) as contrast base
- ONE electric accent only (teal / cyan / cold blue / 
  lavender — character dependent)
- Black ink lines everywhere — no soft edges

TYPOGRAPHY SYSTEM (mandatory for every poster):
- TOP: Character name in large torn washi strip — 
  bold kanji, rough brush
- LEFT SIDE: Single large standalone power kanji — 
  拳 / 王 / 怪物 / 影 (character specific)
- RIGHT SIDE: Vertical tategumi text stack — 
  technique or title name
- BOTTOM LEFT: Action phrase in rough brushstroke
- BOTTOM RIGHT: Death/fate kanji — large, woodblock style
- Scattered ofuda (talisman) strips with small seal kanji
- Red stamp seals (hanko style) — 2 to 3 per image
- ALL TEXT: Japanese only — 
  mix brush calligraphy + woodblock print + stamp seal styles
- Vertical tategumi dominant, some horizontal accent lines

EFFECTS STACK (apply all):
- Ink bleed at every torn paper edge
- Halftone dot pattern on skin and shadows
- Motion blur on hair tips and energy effects
- Supernatural eye glow — only warm/cool accent 
  color in image
- Red ink splatter near fists or collar
- Paper scorch/burn effect at poster edges
- Chiaroscuro lighting — one hard source, 
  deep shadow on 50%+ of face

CHARACTER SWAP LOGIC:
[Itadori] → pink-red spiky hair + rage expression + 
clenched fist forward + red palette + 虎杖悠仁 + 拳 + 
不屈の魂 + spiral background symbol

[Sukuna] → split tattooed face + manic grin + 
sharp teeth + red-black palette + teal eye glow + 
呪いの王 + 伏魔御廚子 + torii + oni mask inserts

[Toji/Megumi corrupted] → black hair over face + 
purple glowing eyes + cold smirk + dark purple palette + 
chains + 怪物 + 術師殺し + 逃げ場なし + crack patterns

[Megumi] → dark hair red-streaked bun + ice blue eyes + 
stoic expression + shikigami wolf+crow flanking face + 
black palette + 影の継承者 + 覚悟 + 死 + tomoe spirals

OUTPUT STYLE REFERENCE: Vizuraai editorial manga poster —
not fan art, not wallpaper — treated as premium 
Japanese cultural print. Vagabond (Inoue) x Berserk x 
JJK Akutami inking precision.

NEGATIVE PROMPT: No English text, no Western comic style, 
no soft anime style, no gradient backgrounds, 
no digital glow effects, no clean skin, 
no symmetrical layouts, no Inter/Roboto fonts.

How to use it:
Keep everything the same
Only swap the CHARACTER SWAP LOGIC block for whichever character you're generating
Change the dominant color accordingly
Everything else stays locked 🔒

3653 / 10000

Based on: A battle-worn shonen warrior glares through torn paper layers

Model

Select a model to continue.

Each model exposes different quality and editing capabilities.

Size & Aspect Ratio

Pick the output framing.

Available options change with the selected model.

Select a model to see credit cost. Reference mode adds 1 credit on supported models.

Generate Similar

Ready when you are

Generating your image

Image ready

We could not finish this image

Generating your image

Image ready

We could not finish this image

Send people to the full prompt and image