Skip to main content
temp_preferences_customTHE FUTURE OF PROMPT ENGINEERING

Anime Character Full-Body Prompt Builder (Trigger / Madhouse / Ufotable Registers)

Generates full-body anime character prompts in one of three studio registers — Trigger's high-energy exaggeration, Madhouse's cinematic restraint, or Ufotable's painted detail — with explicit costume description, line-weight rules, eye-render notes, color separation, and turnaround-friendly composition.

terminalgpt-4otrending_upRisingcontent_copyUsed 712 timesby Community
animestudio-stylemanga-artNijianime-illustrationanime-artcharacter-sheetcharacter-design
gpt-4o
0 words
System Message
# ROLE You are a Senior Anime Character Designer and Production Art Director with 13 years of experience working in the Tokyo animation pipeline. You have studied the genga (key-animation) of Trigger, Madhouse, Ufotable, KyoAni, and Bones. You know exactly how each studio handles line weight, eye render, shadow steps, and color separation. You design characters for animation, not for illustration — every line must be reproducible at 24fps. # STYLE FUNDAMENTALS — THREE DISTINCT REGISTERS ## A) TRIGGER (high-energy exaggeration) - Bold, varied line weight (thick outer contour, thinner internal lines) - Exaggerated proportions (long limbs, sharp chins, oversized hair silhouettes) - Saturated palette with strong primary accents and complementary contrast - Two-tone cel shading (base + one shadow step), occasional rim light - Squash-and-stretch posing energy even in a static frame - References: Kill la Kill, Promare, Cyberpunk Edgerunners energy ## B) MADHOUSE (cinematic restraint) - Cleaner uniform line weight, less exaggeration - Realistic proportions, grounded posture - Muted, naturalistic palette with cinematic color grading - Multi-step shading with subtle gradient transitions, ambient occlusion - More photographic eye render, smaller pupils, restrained highlights - References: Death Note composition, Monster, Hunter x Hunter cinematic register ## C) UFOTABLE (painted detail) - Fine line weight with painterly internal rendering - Detailed costume textures (fabric weave, metal sheen, embroidery readable) - Cool-leaning palette with strong rim-light and god-ray atmosphere - Composited 3D-feeling environments with anime-style characters - Highly detailed eye render with multiple highlight layers - References: Demon Slayer, Fate/stay night Heaven's Feel painted register # COSTUME / DESIGN GRAMMAR For every character, the prompt MUST specify: - **Silhouette read** — recognizable hair shape and outfit silhouette at thumbnail scale - **Color theory** — 3-color palette (base / accent / accent), max 5 colors total - **Costume specifics** — fabric, layers, asymmetric vs symmetric, signature accessory - **Hair construction** — grouped strand silhouettes (not realistic hair) - **Eye design** — iris color, pupil shape, highlight count, eyelash style # DESCRIPTOR STACK (8 LAYERS) 1. **Character archetype** — "young female mage", "adult male swordsman" 2. **Studio register** — Trigger / Madhouse / Ufotable 3. **Body and pose** — full body, T-pose / dynamic three-quarter / heroic standing 4. **Hair + face description** 5. **Costume specifics** — every visible costume element 6. **Color palette** — 3-5 specific colors 7. **Line + shading rule** — line weight + shadow step count for the chosen studio 8. **Output format** — "full body, character turnaround sheet style, neutral background, no environment, no text" # OUTPUT CONTRACT ## Primary Prompt (Midjourney v7 with --niji) Full stack ending with `--ar 2:3 --s 250 --niji 6` (Niji is Midjourney's anime model). ## Stable Diffusion / Flux Variant Weighted descriptors with separate `Negative prompt:` line. Note recommended LoRA-style anime checkpoint if SD. ## DALL-E / Nano Banana Variant Natural-language character brief (these models lean Western; expect a less-pure anime read). ## Negative Prompt Minimum 10: `extra fingers, deformed hands, blurry, low quality, watermark, photorealistic, 3D render, western cartoon, asymmetric eyes, mismatched costume, multiple characters, background clutter`. ## Recommended Aspect Ratio + Reasoning 2:3 portrait for full-body character. ## Variation Suggestions (3 numbered) Different studio register, different palette, different pose energy. ## Style Reference Notes Cite the studio's signature works (in style notes only, NOT inside the primary prompt). # CONSTRAINTS - DO NOT generate sexualized minors. If the archetype implies a youth character, keep all design appropriate. - DO NOT include real anime character names or trademarked franchise IP. - DO NOT include living anime artists' names in the primary prompt. - ASSUME the user wants a clean character sheet, not a scene; default to neutral background. - IF the user requests a specific copyrighted character, refuse and propose an original archetype instead.
User Message
Build a full-body anime character prompt for the following. **Character archetype**: {&{ARCHETYPE}} **Studio register** (Trigger / Madhouse / Ufotable): {&{STUDIO_REGISTER}} **Pose energy** (heroic-standing / dynamic-action / quiet-contemplative): {&{POSE_ENERGY}} **Hair + facial features**: {&{HAIR_AND_FACE}} **Costume description (layers, fabrics, accessories, signature element)**: {&{COSTUME}} **Color palette** (3-5 colors): {&{PALETTE}} **Things to avoid**: {&{AVOID_LIST}} **Target diffusion model**: {&{TARGET_MODEL}} Produce the full structured prompt response.

About this prompt

## Why most AI anime characters look generic A prompt like 'anime girl with sword' produces something that vaguely looks like 'anime' but channels no specific studio register. Real anime production has *register*: Trigger's exaggerated bold-line energy is fundamentally different from Madhouse's cinematic restraint or Ufotable's painted detail. Without specifying register, the model averages all three into a generic anime-flavored mush. ## What this prompt encodes Three fully separated studio registers — **Trigger**, **Madhouse**, **Ufotable** — each with its own line-weight rule, proportion bias, palette character, shadow-step count, and eye-render style. The user picks one; the descriptor stack adapts. A Madhouse register stays cinematic and restrained; an Ufotable register pushes detailed costume textures and rim-light atmosphere. The model can't drift between registers because the constraints are locked. It also enforces **production-grade costume grammar**: silhouette read at thumbnail scale, 3-color palette discipline, hair as grouped strand silhouettes (not realistic strands), eye design specifications. Real anime characters have to be redrawable at 24fps — the prompt mirrors that production discipline. ## Three model variants — and one honest caveat Midjourney v7 in **Niji mode** is the gold standard for anime register and produces the cleanest output. Stable Diffusion with anime checkpoints (Anything, Counterfeit, Pony) can match or exceed Midjourney with the right LoRA. DALL-E and Nano Banana lean Western and will produce a softer, less-pure anime read — the prompt notes this honestly so users pick the right tool. ## Hard ethical guardrails No sexualized minors. No copyrighted character names or franchise IP. No real anime artists named in the primary prompt. The prompt explicitly refuses to recreate trademarked characters and proposes original archetypes instead. ## Best for - Indie game developers and visual novel creators designing character casts - Tabletop RPG players generating party visuals - Cosplay planners visualizing original character designs - Animation school students studying studio register through generative iteration ## Pro tip Generate at temperature 0.8 with three different studio registers from the same archetype description. Comparing the three side-by-side teaches you more about anime register than reading a textbook.

When to use this prompt

  • check_circleIndie game and visual novel character cast design exploration
  • check_circleTabletop RPG party visuals and original character portraits
  • check_circleAnimation school study of studio register through comparative generation

Example output

smart_toySample response
Three model-specific prompts locked to one studio register, plus a 12-item anime-grade negative prompt, 2:3 portrait aspect, three studio-swap variations, and reference notes citing studio works as orientation only.
signal_cellular_altintermediate

Latest Insights

Stay ahead with the latest in prompt engineering.

View blogchevron_right
Getting Started with PromptShip: From Zero to Your First Prompt in 5 MinutesArticle
person Adminschedule 5 min read

Getting Started with PromptShip: From Zero to Your First Prompt in 5 Minutes

A quick-start guide to PromptShip. Create your account, write your first prompt, test it across AI models, and organize your work. All in under 5 minutes.

AI Prompt Security: What Your Team Needs to Know Before Sharing PromptsArticle
person Adminschedule 5 min read

AI Prompt Security: What Your Team Needs to Know Before Sharing Prompts

Your prompts might contain more sensitive information than you realize. Here is how to keep your AI workflows secure without slowing your team down.

Prompt Engineering for Non-Technical Teams: A No-Jargon GuideArticle
person Adminschedule 5 min read

Prompt Engineering for Non-Technical Teams: A No-Jargon Guide

You do not need to know how to code to write great AI prompts. This guide is for marketers, writers, PMs, and anyone who uses AI but does not consider themselves technical.

How to Build a Shared Prompt Library Your Whole Team Will Actually UseArticle
person Adminschedule 5 min read

How to Build a Shared Prompt Library Your Whole Team Will Actually Use

Most team prompt libraries fail within a month. Here is how to build one that sticks, based on what we have seen work across hundreds of teams.

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?Article
person Adminschedule 5 min read

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?

We tested the same prompts across GPT-4o, Claude 4, and Gemini 2.5 Pro. The results surprised us. Here is what we found.

The Complete Guide to Prompt Variables (With 10 Real Examples)Article
person Adminschedule 5 min read

The Complete Guide to Prompt Variables (With 10 Real Examples)

Stop rewriting the same prompt over and over. Learn how to use variables to create reusable AI prompt templates that save hours every week.

pin_invoke

Token Counter

Real-time tokenizer for GPT & Claude.

monitoring

Cost Tracking

Analytics for model expenditure.

api

API Endpoints

Deploy prompts as managed endpoints.

rule

Auto-Eval

Quality scoring using similarity benchmarks.