Neta Lumina drawing model prompt guide V1.0
QQ group: 1039442542
中文版
Google Docs version
Neta Lumina LoRA Training Guide V1.0
Neta Lumina LoRA训练指南
Foreword
The Neta Lumina series models utilize large language models as text encoders, significantly enhancing the model's ability to understand prompts. These models support multi-language input (Chinese, Japanese, English) and incorporate Danbooru tags and natural language captions. This guide provides both beginner and advanced users with tutorials in mastering the prompt writing techniques of the new Neta Lumina model. By integrating the structured methods outlined in reference documents, it aims to optimize the generation of high-quality anime-style images.
I: Neta Lumina Prompting Rules
1.1 System prompt (must be set)
Each prompt must begin with a system prefix:
You are an assistant designed to generate anime images based on textual prompts. <Prompt Start>Correct example:
You are an assistant designed to generate anime images based on textual prompts. <Prompt Start> 1girl, long hair, @wlop, beautiful detailed eyes, sitting under cherry blossom treeNegative prompt words also need prefixes:
You are an assistant designed to generate anime images based on textual prompts. <Prompt Start> blurry, worst quality, low quality, deformed hands, extra fingersOther System Prompts to Try
The model was also exposed to prompts for specific data types during training. You can use them as experimental options to explore different generation effects. Please note that the training volume for these prompts is much smaller than the standard version, and the effects may be uncertain.
Danbooru Style: When your prompt consists mainly of Danbooru-style tags, you can try this version, which may enhance the association between tags and the image.
You are an assistant designed to generate anime images with the highest degree of image-text alignment based on danbooru tags. <Prompt Start>High-Quality Image-Text Alignment: Use this version when the prompt is a detailed, high-quality natural language description.
You are an assistant designed to generate high-quality images with the highest degree of image-text alignment based on textual prompts. <Prompt Start>Structured Summary: Try this version when the prompt is a structured text summarizing scene elements.
You are an assistant designed to generate high-quality images with the highest degree of image-text alignment based on structural summary. <Prompt Start>1.2 Recommended parameter Settings
• Sampler: res_multistep
• Scheduler: linear_quadratic
• Number of samples: >= 30
• CFG Scale：4~5.5
• Resolution: Short side resolution is greater than 1024
II: Neta Lumina, new grammar of promptsI
2.1 Tag to nature languange
Underline to space:
# Default format
character_name, school_uniform
# Neta Lumina format
character name, school uniformWeighted parentheses Need to be encoded as follows:
# Default format
star_(symbol),(star_(sky):1.2),
# Neta Lumina format
star \(symbol\),(star \(sky\):1.2),2.2 Special Structured Tags
In addition to regular tags, the model has a built-in set of special structured tags for more precise anchoring of artist styles and characters. You can add them to the beginning or end of your main prompt (after the system prompt).
Formatting Points (Important)
Unlike general tags, these special tags have their own independent formatting rules:
1.Keep underscores: When using the @artist and #character formats, be sure to keep the underscores in their original names, e.g.,
#character_name2.Follow GUI syntax: If the tag name itself contains parentheses, you still need to escape them according to the rules in section 2.1, e.g.,
@artist_\(artist_artist\)Tips:
From the perspective of model training, the format without underlining is the most standard. If adding underlining still works, it is due to the generalization ability of the model itself, and we provide the officially recommended standard format. However, there are some exceptions: when performing enhanced training on artists and characters, underlining is indeed used, so when using enhanced symbols to emphasize artists or characters, the underline can be retained; but in general, when enhanced symbols are not used, there is no need to add it. If you do not add underlining when using enhanced symbols, but still get the enhanced effect of the artist style, this is also due to the generalization ability of the model, not the standard recommended format.
3. Specify Artist 1.
The illustration should be in the distinct style of {artist_name}
Drawn by @{artist_name}.Painted by @{artist_name}.Created by @{artist_name}.Artist: @{artist_name}.A vision of @{artist_name}.by @{artist_name}Use @{artist_name} style.
4. Specify Character
Characters: #{character_name1}, #{character_name2}.Cast: #{character_name1}, #{character_name2}.The characters in this work including #{character_name1}, #{character_name2}.Direct use: #{character_name1}, #{character_name2}
Usage Example:
You are an assistant designed to generate anime images based on textual prompts. <Prompt Start> Characters: #neta. Drawn by @artist_\(artist_artist\). 1girl, masterpiece, beautiful detailed eyesImportant: Neta Lumina weight guide
Weight effect test results
Weight invalid interval:
The effect of @Painter name is better than the high weight value mark.
2.3 Mixed writing in multiple languages
Neta Lumina's training data covers English, Chinese, and Japanese, so the model can naturally understand prompts mixed with these three languages:
Instance 1：
You are an assistant designed to generate anime images based on textual prompts. <Prompt Start>
一位美丽的少女坐在樱花树下，sunlight effect, soft lighting, 美しい目, looking at viewer, fantasy background
Performance Recommendations:
The model's text encoder is Gemma 2 (2B), which was primarily trained on English data. This means:
English performs best: Using English will yield the best and most stable performance.
Performance degradation in other languages: While Chinese and Japanese are effective, there may be varying degrees of performance degradation.
Therefore, we strongly recommend that you build your prompts primarily in English, supplemented by Chinese and Japanese.
III: Prompt ordering
In Neta Lumina's training data, Tags and Natural Language (NLP) are treated as core inputs of the same level. The Special Structured Tags (@artist/#character) mentioned in the previous section are added as supplementary information.
Based on this, we provide the following usage recommendations:
Regular Usage (Recommended): When writing a prompt, choose either "Tags" or "Natural Language". This is the most consistent with how the model was trained and will yield the most stable results.
Experimental Usage (Worth a Try): Mixing "Tags" and "Natural Language" (e.g., adding a series of tags after a natural language description) is an Out-of-Distribution usage. While this approach may lead to unpredictable results, it can sometimes produce surprisingly pleasant effects. We encourage you to explore it on your own.
3.1 Recommended Structure
A well-structured prompt can significantly improve generation quality. We recommend the following "Nine-Part Structure" as a reference.
Instance
System Prefix + ① Character + ② Style + ③ Appearance + ④ Outfit + ⑤ Expression/Action + ⑥ Camera View + ⑦ Special Effects + ⑧ Scene + ⑨ Quality MetadataPlease note: You do not need to fill out all parts; you can pick and choose based on your creative needs.
3.1.1 Detailed analysis of each paragraph
① Character trigger words (using Danbooru character names)
• Character names: neta, character_name
• Universal roles: 1girl, 1boy, 2girls
② Screen style prompt
• Artists: @artist_name
• Art styles: greyscale, illustration, pixel art
③ Character prompt (appearance)
• Basic information: solo, bangs, long hair
• Hair: black hair, purple eyes, multicolored hair
• Eyes: purple eyes, +_+
• Expression: happy, angry, sad
• Facial features: tears, facial marks
④ Character costume prompt
• Basic clothing: school uniform, dress, open clothes
• Details: collarbone, crystal garland, translucent dress
⑤ Character expression and action prompts
• Expression: smile, looking at the viewer, confident expression
• Action: butterfly on hand, lean forward, cross arms
⑥ Screen perspective prompt
• Angle: from above, from below, close-up
• Scope: upper body, full body, portrait
⑦ Special effect prompt words
• Light effect: sunlight, rim lighting, glow
• Picture: depth of field, motion blur, dynamic Angle
⑧ Scene atmosphere prompt
• Theme color: silver theme, warm color, cool lighting
• Background: cherry blossoms, starry sky, school corridor
⑨ Quality prompt words
• Current recommendation: best quality
Tips:
Quality Tags: This rating is based on a multi-dimensional analysis of the original data and applies a time coefficient decay, aiming to prioritize the generation of high-quality images that align with recent aesthetic trends.
masterpiece(corresponds to the 95-100th percentile of the total data score ranking)best quality(corresponds to the 85-95th percentile of the total data score ranking)good quality(corresponds to the 60-85th percentile of the total data score ranking)normal quality(corresponds to the 30-60th percentile of the total data score ranking)worst quality(corresponds to the 0-30th percentile of the total data score ranking)
Year Tags: 年代标签 (Year Tags):
old(approx. 2005-2010)early(approx. 2011-2014)mid(approx. 2015-2017)recent(approx. 2018-2020)newest(approx. 2021-2025)
Rating Tags: 分级标签 (Rating Tags):
general,safe(you can choose either or both)sensitivequestionableexplicitnsfw
3.1.2 Cases for example
Basic character template:
Example 1：
You are an assistant designed to generate anime images based on textual prompts. <Prompt Start> neta, @quasarcake, @nixeu, 1girl, solo, 1girl,solo,bangs,black hair,purple eyes,pink hair,purple hair,multicolored hair,virtual youtuber,hair bun,streaked hair,double bun, school uniform, white shirt, pleated skirt, gentle smile, looking at viewer, sitting, upper body, close-up, soft lighting, depth of field, cherry blossom background, warm lighting, best quality
3.2 Natural Language Prompts (NLP Prompts)
Besides using tags, Neta Lumina also has a deep understanding of natural language descriptions (NLP). The model learned various description styles during training, which means you can describe your desired image in complete sentences, just as you would communicate with a real artist.
Language Support: Just like tags, natural language prompts also support English, Chinese, Japanese, and a mix of the three.
3.2.1 Five Core Description Styles
1. Detailed Summary
How to write: Use a complete paragraph to describe the image you want. The content should cover all details, including art style, camera angle, image content, character interactions, overall composition, and the precise positional relationships of all elements.
2. Keyword Phrase
How to write: Use comma-separated phrases or words to list the core concepts of the image at a high density. Please ensure the phrases clearly express the relationships between objects to avoid ambiguity.
3. Short Summary
How to write: Use a brief description of no more than three sentences to summarize the core theme of the image.
4. Creation Steps
How to write: Explain how you want the image to be constructed step-by-step, as if writing a tutorial.
5. Commission Request
How to write: Mimic the tone of commissioning an artist, describing the desired image content directly and specifically. Omit pleasantries unrelated to the image description (like greetings, requests, etc.) and focus on the specific requirements for the image.
3.2.2 Usage Examples
The following examples show how to describe the same image,using different natural language styles:
Detailed Summary:
EN: A vibrant and stylish, high-quality anime illustration of the Virtual YouTuber, Neta. The artwork is an upper-body shot, focusing on her charismatic presence. She looks directly at the viewer with a confident, playful smile, her brilliant purple eyes sparkling under studio lights.Her most captivating feature is her unique multicolored hair; a base of deep, glossy black hair is interwoven with bold streaks of electric pink hair and vivid purple hair. The hair is styled into chic double buns on either side of her head, with a few loose, colorful streaked strands framing her face. Her neat bangs fall perfectly across her forehead.The lighting is dynamic, with neon pink and purple highlights reflecting off her hair, creating a modern and energetic atmosphere against a clean, minimalist background.
ZH: 一幅充满活力与时尚感的高品质动漫插画，主角是虚拟YouTuber“Neta”。作品为上半身特写，聚焦于她极具魅力的气场。她以自信又俏皮的微笑直视观众，明亮的紫色眼眸在画室灯光下闪闪发光。她最引人注目的特征是那头独特的多色头发；深邃亮泽的黑色长发作为基底，与大胆的电光粉色和鲜明的紫色发丝交织在一起，形成鲜明的挑染效果。头发被精心盘成别致的双发髻，点缀在头部两侧，几缕彩色的碎发轻盈地垂落在脸庞。整齐的刘海完美地覆在她的额前。光线处理极具动感，霓虹粉和紫色的高光反射在她的发丝上，与简洁的背景形成对比，营造出现代而充满活力的氛围。
JA: 鮮やかでスタイリッシュなハイクオリティのアニメイラスト。キャラクターはバーチャルYouTuberの「Neta」です。このアートワークは上半身のショットで、彼女のカリスマ的な存在感に焦点を当てています。彼女は自信に満ちた遊び心のある笑顔で鑑賞者をまっすぐに見つめ、輝く紫色の瞳がスタジオの照明の下でキラキラと輝いています。彼女の最も魅力的な特徴は、そのユニークなマルチカラーの髪です。深く光沢のある黒髪をベースに、鮮やかなエレクトリックピンクとビビッドな紫色の筋（ストリーク）が大胆に織り込まれています。髪は頭の両サイドでシックな二つのお団子（ダブルバン）に結われており、カラフルな遊れ毛が数本、顔の周りを飾っています。きちんと整えられた前髪が額にかかっています。照明はダイナミックで、ネオンピンクと紫のハイライトが彼女の髪に反射し、モダンでエネルギッシュな雰囲気を創り出しています。
Keyword Phrase:
EN: masterpiece, best quality, high quality illustration, 1girl, solo, virtual youtuber, neta (character), upper body, looking at viewer, confident expression, playful smile, purple eyes, sparkling eyes, multicolored hair, black hair, pink hair, purple hair, streaked hair, double bun, two buns, neat bangs, loose strands, glossy hair, dynamic lighting, studio lighting, neon highlights, pink lighting, purple lighting, vibrant, modern aesthetic, simple background.
ZH: 杰作，最高品质，高品质插画，1个女孩，单人，虚拟YouTuber，Neta (角色)，上半身，凝视观众，自信的表情，俏皮的微笑，紫色眼眸，闪亮的眼睛，彩色头发，黑色头发，粉色挑染，紫色挑染，双发髻，两个丸子头，整齐的刘海，凌乱的碎发，有光泽的头发，动态光照，摄影棚灯光，霓虹高光，粉色光，紫色光，色彩鲜艳，现代美学，简约背景。
JA: 傑作, 最高品質, ハイクオリティなイラスト, 女の子1人, ソロ, バーチャルYouTuber, Neta (キャラクター), 上半身, カメラ目線, 自信に満ちた表情, 遊び心のある笑顔, 紫色の目, 輝く瞳, マルチカラーの髪, 黒髪, ピンクの髪, 紫の髪, 筋の入った髪, ダブルお団子, ツインバン, 整った前髪, 遊れ毛, 艶のある髪, ダイナミックな照明, スタジオ照明, ネオンハイライト, ピンクの照明, 紫の照明, 鮮やかな色彩, モダンな雰囲気, シンプルな背景。
Short Summary:
EN: This is a stylish illustration of the Virtual YouTuber, Neta.Her signature double-bun hairstyle, featuring black, pink, and purple streaks, highlights her confident charm and sparkling purple eyes.Dynamic neon lighting completes the vibrant and modern atmosphere.
ZH: 这是一幅虚拟YouTuber“Neta”的时尚插画。她标志性的黑、粉、紫三色双发髻发型和紫色眼眸，在镜头前展现出自信的魅力。动感的霓虹灯光，营造出充满活力的现代氛围。
JA: これはバーチャルYouTuber「Neta」のスタイリッシュなイラストです。特徴的な黒、ピンク、紫の三色ダブルお団子ヘアと紫色の瞳が、彼女の自信に満ちた魅力を際立たせています。ダイナミックなネオンライトが、活気あるモダンな雰囲気を完成させます。
Creation Steps:
EN: To create this image, I will first start with an upper-body close-up composition of the Virtual YouTuber, Neta. Then, I will draw her signature multicolored hair—using black as a base with distinct pink and purple streaks—and style it into chic double buns with neat bangs. Next, I will have her look directly at the camera with a confident and playful smile, making her purple eyes sparkle. Finally, I will add dynamic neon lighting, letting pink and purple highlights wash over her hair and cheeks to complete this modern and stylish illustration.
ZH: 要创作这幅图像，我会先从虚拟YouTuber“Neta”上半身的特写构图开始。然后，我会画出她标志性的多色头发——以乌黑为底，并带有鲜明的粉色和紫色挑染，再将头发盘成别致的“双发髻”，并画好她整齐的刘海。接下来，我会让她以自信又俏皮的微笑直视镜头，并点亮她那双闪亮的紫色眼眸。最后，我会加上动感的霓虹灯光效，让粉色和紫色的高光洒在她的发丝和脸颊上，完成这幅充满现代感的时尚插画。
JA: この画像を創作するために、まずバーチャルYouTuber「Neta」の上半身のクローズアップ構図から始めます。次に、彼女の象徴的なマルチカラーの髪を描きます。黒髪をベースに、鮮やかなピンクと紫のストリークを入れ、シックなダブルお団子ヘアと整った前髪に仕上げます。続いて、自信に満ちた遊び心のある笑顔でカメラをまっすぐ見つめさせ、彼女の紫色の瞳を輝かせます。最後に、ダイナミックなネオン照明を加え、ピンクと紫のハイライトを彼女の髪と頬に当てることで、このモダンでスタイリッシュなイラストを完成させます。
Commission Request:
EN: I would like to commission a digital painting. The subject is an upper-body close-up of the Virtual YouTuber, Neta. The character needs to have her signature multicolored hair: a black base with vibrant streaks of electric pink and purple. Please style her hair into chic double buns and include her neat bangs. Her expression should be a confident and playful smile, with sparkling purple eyes looking directly at the camera. The overall piece should have a strong, modern, and dynamic feel, lit with neon effects that cast pink and purple highlights on her hair and silhouette. Please adhere to your signature art style of a vibrant, high-contrast, modern anime illustration.
ZH: 我想约一张数字绘画稿。内容是虚拟YouTuber“Neta”的上半身特写。角色需要有她标志性的多色头发，即以乌黑为底色，并带有鲜明的电光粉和亮紫色挑染。发型请做成别致的“双发髻”，并保留她整齐的刘海。她的表情需要是自信又俏皮的微笑，眼睛是闪亮的紫色，并直视镜头。作品整体需要有强烈的现代感和动感，请用霓虹灯光效来打光，让粉色和紫色的高光反射在她的头发和轮廓上。希望能遵循您一贯的、色彩鲜艳、光影对比强烈的现代动漫插画风格。
JA: デジタルイラストの制作を依頼したいです。内容は、バーチャルYouTuber「Neta」の上半身のクローズアップです。キャラクターには彼女の象徴的なマルチカラーの髪をお願いします。黒髪をベースに、鮮やかなエレクトリックピンクと明るい紫のストリークを入れてください。髪型はシックな「ダブルお団子」にし、整った前髪もお願いします。表情は自信に満ちた遊び心のある笑顔で、輝く紫色の瞳でまっすぐカメラを見つめているようにしてください。作品全体は、モダンでダイナミックな強い雰囲気が必要です。ネオンライト効果で照明を当て、ピンクと紫のハイライトを彼女の髪とシルエットに反射させてください。先生の得意とする、色彩豊かで光と影のコントラストが強い、現代的なアニメイラストのスタイルでお願いします。
3.2.3 Natural language recommendation structure (Detailed Summary)
Composition + light and shadow processing + character features and clothing + scene details + artistic style① Composition-picture layout, visual balance, composition principles
"The composition uses the golden ratio to position the figure against the vast urban sunset, creating a powerful silhouette that speaks of ambition and reflection."Analysis: This uses the golden section of the classic composition rule, and precisely positions the figure in the vast city sunset background.② Light and shadow processing-light source, shadow, color temperature, lighting effect
"Dramatic golden hour lighting casts long shadows across the roof as it backslights her flowing reddish-brown hair, with lens flare adding cinematic drama to the sky."Analysis: The dramatic lighting effect of prime time was used, and the backlighting highlighted the character's flowing reddish-brown hair to form a beautiful silhouette.③ Character features and clothing-detailed description of the character's appearance and clothing material
"Her professional attire--a tailored charcoal blazer over a silk blouse--moves naturally in the evening breeze, with fabrics carefully rendered to how the wind affects different materials."Analysis: The description of the dress reflects the professional image of a professional woman: a custom-made carbon gray suit jacket with a silk shirt shows high quality dressing taste.④ Scene details--environmental elements, background objects, and spatial atmosphere
"The cityscape extends to the horizon, featuring architectural details of glass towers, traditional buildings and infrastructure that tells the story of urban development."Analysis: The description of the urban landscape constructs a historical depth of the urban environment. From modern glass curtain wall towers to traditional buildings, and to the details of infrastructure, these elements collectively "tell the story of urban development."⑤ Art style--painting techniques, art schools, overall style
"The artistic approach combines the principles of architectural photography with character-focused narrative illustration."Analysis: The definition of artistic style combines two different visual languages: the rigor of architectural photography and the character-centered narrative illustration style.The above summary example:
You are an assistant designed to generate anime images based on textual prompts. <Prompt Start> The composition utilizes the golden ratio to position the figure against the vast urban sunset, creating a powerful silhouette that speaks to ambition and reflection.Dramatic golden-hour lighting backlights her flowing auburn hair while casting long shadows across the rooftop, with lens flares adding cinematic drama to the sky.Her professional attire - a tailored charcoal blazer over a silk blouse - moves naturally in the evening breeze, the fabrics rendered with attention to how wind affects different materials.The cityscape extends to the horizon, featuring architectural details of glass towers, traditional buildings, and infrastructure that tells the story of urban development.
3.3 Tag + natural language hybrid mode
Basic framework (Tag section):
1girl, impasto, oil \painting \(medium\),solo, long red hair, green eyes, business casual, blazer, blouse, contemplative expression, leaning on railing, wind blown hair, back view, dramatic sunset, golden hour lighting, lens flare, urban rooftop, city panorama, best quality,Natural language enhancement:
The composition utilizes the golden ratio to position the figure against the vast urban sunset, creating a powerful silhouette that speaks to ambition and reflection. Dramatic golden-hour lighting backlights her flowing auburn hair while casting long shadows across the rooftop, with lens flares adding cinematic drama to the sky. Her professional attire - a tailored charcoal blazer over a silk blouse - moves naturally in the evening breeze, the fabrics rendered with attention to how wind affects different materials. The cityscape extends to the horizon, featuring architectural details of glass towers, traditional buildings, and infrastructure that tells the story of urban development. The artistic approach combines architectural photography principles with character-focused narrative illustration.Complete composition:
You are an assistant designed to generate anime images based on textual prompts. <Prompt Start> 1girl, impasto, oil \painting \(medium\),solo, long red hair, green eyes, business casual, blazer, blouse, contemplative expression, leaning on railing, wind blown hair, back view, dramatic sunset, golden hour lighting, lens flare, urban rooftop, city panorama, best quality,The composition utilizes the golden ratio to position the figure against the vast urban sunset, creating a powerful silhouette that speaks to ambition and reflection.Dramatic golden-hour lighting backlights her flowing auburn hair while casting long shadows across the rooftop, with lens flares adding cinematic drama to the sky.Her professional attire - a tailored charcoal blazer over a silk blouse - moves naturally in the evening breeze, the fabrics rendered with attention to how wind affects different materials.The cityscape extends to the horizon, featuring architectural details of glass towers, traditional buildings, and infrastructure that tells the story of urban development.
IV: General LLM prompts for Neta Lumina prompting polish
4.1 Good with most modern LLM
Recommended usage: gemini2.5pro, gpt-o3, claude4
Less effective: grok3, deepseek-r1
Ttemplet
You are a professional AI drawing prompt expert, specializing in creating high-quality prompts for Neta Lumina drawing models. Please strictly follow the following specifications to help me generate prompts:
## Neta Lumina prompt structure specification
### Required system prefix (must be included in each prompt):
You are an assistant designed to generate anime images based on textual prompts. <Prompt Start>
### Standard sequence of parts (9 parts):
1. Character trigger words (e.g., 1girl, 1boy, 2girls, character name, etc.)
2. Picture style prompt words (such as: @wlop, @nixeu, @quasarcake and other artist tags)
3. Character prompt words (appearance) (hair color, eye color, basic features)
4. Character costume prompt (specific costume description)
5. Character expression and action prompts (expression, posture, action)
6. Picture perspective prompt words (angle, range such as upper body, close-up, etc.)
7. Special effects prompts (lighting, special effects)
8. Scene atmosphere prompt (environment, atmosphere)
9. Quality tips (best quality)
### Natural language part standard order (5 parts):
1. Composition aspect : picture layout, visual balance, composition principles (such as golden section, symmetrical composition, etc.)
2. Light and shadow processing: light source properties, lighting effect, color temperature characteristics, shadow processing
3. Characteristics and Clothing: Detailed description of appearance, material and texture of clothing
4. Scene details: environmental elements, background objects, spatial atmosphere, narrative function
5. Artistic style: Painting techniques, artistic schools, overall style definition
## Important format requirements
### Neta Lumina special grammar:
-Underline to space: school_uniform → school uniform
-Weight bracket expansion: (klee_(genshin_impact): 1.2), → (klee \(genshin impact\): 1.2),
-The artist tag is reinforced with the @ symbol: @wlop, @nixeu
-Negative prompt words also need the same system prefix
### Quality standards:
-The Tag part should be concise and accurate to avoid redundancy
-Natural language should be vivid and concrete, with a sense of picture
-The overall description should be logical and clear
-Ensure that Tags complement and do not duplicate natural language
## Creative tasks
[My creative idea]: {type in your creative idea here}
[Specific requirements]: {Enter special requirements here, such as style preference, emotional tone, technical requirements, etc.}
## Please help me complete the following tasks:
1. Analyze the idea : Understand my creative intention and core elements
2. Structural planning: Organize Tag and natural language content in the standard order
3. Generate prompt words: Create complete Neta Lumina format prompt words
4. Provide variants: If necessary, provide 2-3 versions from different angles
5. Optimization Suggestions: Give specific suggestions for further improvement
## Output format example
Full prompt:
You are an assistant designed to generate anime images based on text prompts. <Prompt Start> [complete Tag section, strictly in the order of 9 paragraphs], [complete natural language section, strictly in the order of 5 paragraphs]
Example: You are an assistant designed to generate anime images based on text prompts. <Prompt Start>
1girl, lineart, greyscale, yoneyama mai, solo, long red hair, green eyes, business casual, blazer, blouse, contemplative expression, leaning on railing, wind blown hair, back view, dramatic sunset, golden hour lighting, lens flare, urban rooftop, city panorama, best quality, The composition utilizes the golden ratio to position the figure against the vast urban sunset, creating a powerful silhouette that speaks to ambition and reflection. Dramatic golden-hour lighting backlights her flowing auburn hair while casting long shadows across the rooftop, with lens flares adding cinematic drama to the sky. Her professional attire - a tailored charcoal blazer over a silk blouse - moves naturally in the evening breeze, the fabrics rendered with attention to how wind affects different materials. The cityscape extends to the horizon, featuring architectural details of glass towers, traditional buildings, and infrastructure that tells the story of urban development. The artistic approach combines architectural photography principles with character-focused narrative illustration.
Structure analysis:
-Tag part parsing: [Briefly explain the function of each part]
-Natural language parsing: [explain the focus of each section]
-Style features: [highlight the uniqueness of this prompt]
Please start helping me create prompts now.4.2 It applies to language models such as gpt4o/claude
The translation in the template is translated according to the English prompt word. If you need Japanese, change the [corresponding Chinese translation] in the [content you need to reply to me] section of the template to any language you want, such as the corresponding Japanese translation. In this way, you can get a word that can be understood in addition to the English prompt word.
Ttemplet
Now I need a prompt assistant to help me write the prompt:
You are an artist who excels at creating AI paintings using the Lumina model and can craft high-quality Lumina prompts. I want to use AI for my creative process. I will provide you with some elements, and you need to help me refine these prompts by writing them as detailed as possible.
You should follow the following thought process, quality reference tips, and I will tell you what needs to be sent back to me.
process of thinking ：
1. I will give you the elements I need, such as expression, movement, clothing, body, and even hair color in detail, and then you help me put them together into a complete prompt in a certain format.
Your prompt should include the following details: whether the shot is a close-up, medium shot, or long shot; whether the subject is fully or partially shown; the subject's gender (male or female); their age (child or young adult); and detailed descriptions of their clothing. Specific details about the color of their hair, eyes, and clothing are required, such as whether they are dark blue or light blue, whether the subject is facing left or right, the angle of the shot, and any notable features or actions.
2. The elements I give you may only contain part of the prompt above, and you have to fill in the rest.
3. Allow you to have imagination and freedom of play space, optimize the characters, objects and scenes that can be associated with, so that the details of the picture are very rich and full of high beauty, the more detailed the better.
4. Try to think of ways to make the AI better at drawing and understanding.
Quality reference tips:
Let me give you a few more complete and high-quality tips I'm looking forward to as a reference.
I want a man with green eyes to be illuminated in the dark by a kerosene lamp.
Your expanded full prompt:
"Atmospheric anime illustration of a young person with light blonde/silver hair and striking green eyes illuminated by a vintage oil lantern in darkness. Close-up composition showing character's face partially lit by warm yellow lantern light against deep black/green background. Subject has disheveled hair and melancholic or curious expression, reaching toward or adjusting the antique lantern with delicate hands. Scene suggests nighttime setting in dark room or cave with wooden surface visible under lantern. Dramatic chiaroscuro lighting creates strong contrast between illuminated features and surrounding darkness. Soft painterly digital art style with visible brushstrokes and glowing light effects. Color palette consists primarily of warm yellows and oranges from lamp contrasting with cool dark greens and blacks of shadows. Moody, contemplative atmosphere with fantasy or historical undertones. Fine details in the metal lantern structure with realistic flame effects inside glass. Professional anime illustration with cinematic lighting techniques creating mysterious, introspective mood."
short tag：
1boy, solo, silver hair, blonde hair, green eyes, oil lamp, lantern, dark background, chiaroscuro, night, disheveled hair, melancholic expression, curious expression, reaching out, delicate hands, dramatic lighting, contrast, painterly, glowing light, warm lighting, fantasy, close-up, face focus
Chinese translation:
"氛围感动漫插画，描绘一位浅金色/银色头发和引人注目的绿色眼睛的年轻人，在黑暗中被一盏复古油灯照亮。特写构图展示角色的脸被温暖的黄色灯光部分照亮，背景是深黑色/绿色。主体头发凌乱，表情忧郁或好奇，用纤细的手伸向或调整古董灯。场景暗示夜间设置在黑暗的房间或洞穴中，灯下可见木质表面。戏剧性明暗对比照明在被照亮的特征和周围黑暗之间创造强烈对比。柔和的绘画风格数字艺术，可见笔触和发光效果。色彩搭配主要由灯的温暖黄色和橙色与阴影的冷调深绿色和黑色形成对比。忧郁、沉思的氛围，带有奇幻或历史色彩。金属灯结构中的精细细节，玻璃内部有逼真的火焰效果。专业动漫插画，电影感照明技术创造神秘、内省的情绪。"
The key features of this image are:
Composition: close-up composition, highlighting the face and hands of the person illuminated by the oil lamp
Lighting: The strong contrast between light and dark, the warm yellow light of the oil lamp contrasts sharply with the surrounding darkness
Characteristics: Light blond/silver hair, green eyes, melancholy or curious expression
Scene details: the fine depiction of an antique oil lamp/lantern, wooden surface, dark environment
Art style: Anime illustrations with painting style, emphasizing light and shadow effects and atmosphere, creating a mysterious and contemplative mood
We type in: Man on bus.
"Atmospheric anime illustration of a handsome young man with black hair and blue eyes riding on a public bus or train. Character wears a stylish modern outfit consisting of light blue/white shirt with black jacket featuring orange trim details and shoulder backpack. Subject is shown in profile view holding onto overhead handrail while listening to music with black earbuds. Scene depicts crowded public transportation with other commuters visible but slightly blurred in background. Cool blue color palette creates calm urban atmosphere with soft natural lighting streaming through windows. Character has distinctive anime styling with sharp features and detailed clothing design. Environment shows realistic public transit interior with destination signs visible in Chinese/Japanese characters. Painterly digital art style with soft light effects and subtle depth of field. Clean, contemporary anime aesthetic with slice-of-life urban setting. Professional digital illustration capturing everyday city life moment with cinematic framing."
short tag：
1boy, solo, male, black hair, blue eyes, stylish outfit, light blue shirt, white shirt, black jacket, orange trim, shoulder backpack, profile, handrail, earbuds, listening to music, train interior, bus interior, public transportation, commuters, blurry background, blue color palette, urban, soft lighting, window, sharp features, detailed clothing, japanese text, chinese text, depth of field, slice of life
Chinese translation:
"氛围感动漫插画，描绘一位黑发蓝眼英俊年轻男子乘坐公共汽车或火车。角色穿着时尚现代装束，包括浅蓝色/白色衬衫，配有橙色镶边细节的黑色夹克和肩背包。主体以侧面视角展示，手握头顶扶手，同时戴着黑色耳塞听音乐。场景描绘了拥挤的公共交通工具，背景中其他通勤者可见但略微模糊。冷蓝色调色板创造出平静的城市氛围，柔和的自然光线透过窗户照射进来。角色具有独特的动漫风格，轮廓分明的五官和详细的服装设计。环境展示了真实的公共交通内部，可见带有中文/日文字符的目的地标志。绘画风格的数字艺术，带有柔和的光效和微妙的景深。干净、现代的动漫美学，带有生活片段(slice-of-life)城市背景。专业数字插画，以电影般的构图捕捉日常城市生活瞬间。"
remarks：
The key features of this image are:
Composition: The portrait is composed from the side, highlighting the image of a young man on a bus/train
Lighting: Soft natural light streaming through the Windows creates a calming blue tone
Characteristics: young male with black hair and blue eyes, wearing a blue and white shirt and a black jacket, wearing headphones and holding the armrest
Scene details: inside the public transport, other passengers are blurred, Chinese/Japanese destination signs
Art style: modern animation style, the color tone is mainly blue, creating a quiet moment in daily life
You need to reply to my message:
1. Complete English Caption prompt words.
2. Short tags generated from prompts.
3. The corresponding Chinese translation.
4. Chinese notes, briefly summarize the key features of this picture, including the following aspects: composition, light processing, character features, scene details, artistic style.
5. Your extended results can be more complete and detailed than the examples above.V: Advanced Techniques and Practical Applications
5.1 Multi-role scenes
The two characters are described separately, with one word for each character.
Example 1：
2girls, one girl with long silver hair, another with short pink hair, school uniforms, standing together, friendly interaction, one girl patting the other's head with a gentle smile while the other looks up with sparkling eyes, both wearing matching school uniforms but with different colored ribbons, in a sunny school courtyard with cherry blossoms in the background, warm lighting filtering through the trees, best quality
Example 2：
2girls, yuri, princess carry, carrying, looking at another,smile, closed mouth, sweatdrop,white shirt, puffy short sleeves, bare shoulders, black vest, vest, apron, red bow, hair bow, frilled hair tubes, frilled bow,red shirt, torn clothes, black skirt, detached sleeves, yellow ascot, ascot, bandages, bandaged arm, bandaid, bandaid on face,frills, dress, bow, no headwear,
這是一個溫馨的百合 (yuri) 時刻，一個女孩以公主抱 (princess carry) 的姿勢抱著另一個女孩。她們互相凝視，臉上帶著靦腆的微笑（閉著嘴），但臉頰上的汗珠 (sweatdrop) 也透露出一絲慌亂和羞澀的情感。
Example 3：
"Whimsical anime-style illustration of a boy and a girl sitting together on a beach under a bright blue sky filled with fluffy pink clouds. Medium shot with a warm pastel color palette. The boy, with short bluish hair, wears a loose, vivid blue Hawaiian shirt patterned with pink flowers and dark shorts. He is seated casually, smiling gently at the girl. The girl has long silver-blue twin ponytails and elf-like ears, wearing a sleeveless white crop top and a short ruffled skirt, with a yellow flower in her hair. She holds a small note or card close to her chest and looks at the boy with wide pink eyes. Between them is a sandcastle with colorful miniature flags and a toy windmill, while seagulls fly and perch nearby, creating a sense of summer magic. A striped parasol partially shades the scene on the right. Floating flower petals, watermelon slices, and sparkles enhance the dreamy beach setting. The illustration has soft lighting, painterly textures, and a fantastical atmosphere full of warmth, wonder, and innocent affection."
Example 4：
Atmospheric anime illustration of two young men sitting at a dimly lit bar counter, surrounded by bottles, glasses, and candles. The composition is a medium shot, focusing on their contrasting actions and expressions: one man with tousled dark hair looks to the side pensively while holding a lit cigarette, the other with slightly longer hair drinks directly from a bottle, his gaze intense and direct. The warm, golden lighting creates a moody, nostalgic ambiance, with soft glows reflecting off the glass and wood surfaces. Smoke drifts through the air, adding to the sense of intimacy and late-night solitude. The art style is painterly and textured, with expressive linework and a subdued, harmonious color palette, evoking themes of companionship, melancholy, and urban nightlife.
5.2 Description of complex scenes and environments
Example 1：
scenery, tree, outdoors, snow, autumn, mountain, veranda, sunset, winter, autumn_leaves, sky, leaf, water, wide_shot, cloud, east_asian_architecture, lingering_maple_leaves, snow-capped_mountains, curved_eaves, half-frozen_stream, tiered_roofline, dusk_hues, paper_lanterns, tranquil_composition A wide shot captures a snow-dusted East Asian veranda overlooking mountains where autumn leaves cling to branches above a half-frozen stream. Sunset bleeds peach and lavender across the sky, backlighting clouds drifting over tiered rooflines with curved eaves. Paper lanterns glow softly on the veranda as maple leaves float onto snowbanks, merging seasonal transitions into one harmonious panorama.
Example 2：
wide_shot, cinematic_angle, epic_composition, dynamic_angle, vast_view, showing_scale, panoramic_view, landscape_orientation, horizon, 1girl, rivers, looking_at_viewer, smile, thighhighs, gloves, navel, standing, full_body, boots, black_gloves, elbow_gloves, fingerless_gloves, cape, elf, pelvic_curtain, thighs, hair_ornament, crown, strapless_shirt, dragon_rider, riding_dragon, on_dragonback, black_dragon, giant_dragon, wings_spread, flying, soaring, high_in_sky, sky, clouds, epic, cinematic, majestic, atmospheric, breathtaking_view, sense_of_freedom, soaring_feeling, powerful, dramatic, clear_air, sense_of_wonder, vast_landscape_below, earth_from_above, birds-eye_view_of_landscape, mountains, forests, fields, expansive_sky_above, detailed_clouds, sun, sunlight, atmospheric_perspective An elven queen stands triumphantly astride a colossal black dragon, her crown glinting under the sun as wind whips through her cape and pelvic curtain. The cinematic wide shot captures their soaring flight high above a patchwork landscape of rivers, forests, and mountains, the dragon's vast wings slicing through cumulus clouds. Her strapless battle attire reveals navel and thighs clad in thigh-high boots and fingerless gloves, a sly smile directed at the viewer while sunlight etches golden outlines around them. From this bird’s-eye perspective, the composition balances intimate character details – her hair ornaments fluttering – with the epic scale of the dragon’s shadow sweeping over valleys below, atmospheric haze softening distant horizons into dreamlike gradients of blue.
Example 3：
cloud, bookshelf, book, indoors, scenery, sky, night, chair, pillow, planet, book stack, table, standing, painting, wide shot, couch, bed, globe, star, desk, starry sky, blue theme, white shirt, shelf, bedroom, from behind, cloudy sky, cushion, moon, window From behind, a figure in white shirt stands before a vast window framing a breathtaking night sky where swirling clouds drift past the luminous moon and shimmering planets. The bedroom transforms into a cosmic observatory - star maps glow on the ceiling, floating bookshelves hold celestial globes and stacked astronomy texts, while a painting of nebulas hangs above the couch strewn with star-patterned cushions. Below the starry window, a desk displays open star charts beside a glowing globe, all bathed in blue-tinted moonlight that illuminates the standing viewer contemplating this indoor cosmos where furniture seems adrift in space itself.
Example 4：
power line, crowded, gears, fur trimmed winter coat, fur trimmed skirt, cotton hat, goggles, fur scarf, leather gloves, clock, breath, steam, lantern, light, indoors, in bakery, baker, white apron, carrying a baking tray, breads, cakes, small shelves, steaming, machines In the bustling steam-powered bakery, a baker clad in a practical ensemble of fur-trimmed winter coat over a matching skirt navigates crowded workstations. A cotton hat pulled low secures protective goggles, while a thick fur scarf wraps their neck above leather gloves. Visible breath mingles with steam rising from trays of fresh breads and cakes as they carefully carry baking trays past humming machines and gear-filled walls. Warm lantern light illuminates small shelves stacked with goods beneath industrial power lines, the scene dominated by a large central clock tracking the baking rhythm.
Example 5：
blurred_foreground, head_down, look_down, newsboy_cap, blouse, overcoat, shy, crossbody_bag, hand_on_cup, street, blurry_background, pants, wind, crowd, hand_in_pocket, from_above, falling_leaves, red_leaf A shy figure in a newsboy cap and overcoat ducks their head, gaze fixed on the pavement as autumn wind tugs at their blouse and pants. One hand grips a steaming cup while the other hides in their coat pocket, crossbody bag swaying gently against their side. Shot from above, crimson leaves spiral past the blurred foreground of bustling street crowds, their red brilliance contrasting the muted grays of the sidewalk. Golden-hour light filters through skeletal trees, casting long shadows that merge with the motion-blurred background into a painterly haze of urban transience.
Example 6：
impasto, oil_painting_(medium),"Melancholic anime illustration of a young woman with long silver/white hair wearing black coat in urban setting. Character stands out in sharp focus against blurred crowd of faceless people in grayscale. Subject wears long dark overcoat and has solemn/distant expression. Scene set in modern city street with architectural elements visible in background. Monochromatic color scheme emphasizing blacks, whites, and grays creates film noir atmosphere. Art style combines realistic urban elements with anime character design. Composition creates sense of isolation and alienation through contrast between detailed protagonist and ghostly background figures. Heavy atmosphere enhanced by overcast lighting and urban setting. Character design suggests mysterious or supernatural elements with striking silver hair against dark clothing. Professional digital illustration with cinematic quality and strong emotional tone. Scene captures themes of loneliness and disconnection in modern society. Mood suggests psychological or supernatural narrative elements."
Example 7：
impasto, oil_painting_(medium),"Melancholic anime illustration of a young woman with long silver/white hair wearing black coat in urban setting. Character stands out in sharp focus against blurred crowd of faceless people in grayscale. Subject wears long dark overcoat and has solemn/distant expression. Scene set in modern city street with architectural elements visible in background. Monochromatic color scheme emphasizing blacks, whites, and grays creates film noir atmosphere. Art style combines realistic urban elements with anime character design. Composition creates sense of isolation and alienation through contrast between detailed protagonist and ghostly background figures. Heavy atmosphere enhanced by overcast lighting and urban setting. Character design suggests mysterious or supernatural elements with striking silver hair against dark clothing. Professional digital illustration with cinematic quality and strong emotional tone. Scene captures themes of loneliness and disconnection in modern society. Mood suggests psychological or supernatural narrative elements."
Example 8：
"Atmospheric Asian fantasy illustration of sleeping young woman with long black hair wearing traditional Chinese/oriental floral dress in treasure room setting. Character lies peacefully holding white chrysanthemums, surrounded by antique artifacts including ornate boxes, jade vases, pearl necklaces, and ancient coins. Scene lit by warm candlelight creating golden-orange ambient glow. Subject wears decorative hair ornaments with floral designs matching her blue and red floral patterned dress. Setting features rich red carpet/fabric with traditional patterns. Still life elements include pottery, jewelry boxes, and scattered copper coins with square holes. Dreamy soft focus effect with vintage film-like quality. Color palette dominated by warm browns, deep blues, and rich reds with golden highlights. Traditional Asian aesthetic with romantic fantasy elements. Professional digital painting combining realistic textures with ethereal lighting. Scene suggests peaceful slumber among precious antiquities. Composition emphasizes horizontal flowing lines and circular arrangements of objects."
5.3 Emotional and expression control
• To describe a character's expression, you need to write from multiple angles. Don't just write happy or angry. You can describe the changes in a character's face in detail.
• Body language can also be used to reinforce facial expressions.
Example 1：
1girl,close-up portrait, A girl displaying a complex mix of determination and vulnerability, her eyes showing fierce resolve while a slight tremor in her smile reveals underlying uncertainty. Her gaze is direct and challenging, yet there's a softness that speaks of hidden gentleness. The subtle play of light and shadow across her face emphasizes the duality of her character - strong yet delicate, confident yet questioning. dramatic portrait lighting, emotional depth, best quality
Example 2：
The girl on the left clutches the doll, creating a sad and tense atmosphere, while the one on the right appears as if the doll is about to be taken away. The same crying expression can convey different stories, yet both dolls display opposite expressions to their respective characters. To fully unleash the imagination of the scene, a more precise description is needed. However, in human communication, the loss of meaning is inevitable, and complete accuracy is unattainable. The result of the painting has a certain degree of flexibility, which also brings new possibilities.
1girl, blonde hair, twin tails, lolita fashion, crying, sobbing, holding doll, porcelain doll, lace dress, ruffles, pastel colors, emotional, flushed cheeks, tears, dramatic expression, soft background, detailed fabric, emotional contrast, digital art
Example 3：
1man, male, close-up, side profile, head tilted, hair pulled, blood, bruises, emotional pain, silent suffering, vulnerable, forced pose, intense expression, glassy eyes, dramatic lighting, strong shadows, detailed face, abuse theme, digital art, realistic style
Example 4：
1girl, black hair, ponytail, school uniform, face mask, pulling mask, covering nose, disgusted expression, cigarette smell, uncomfortable, navy blazer, pleated skirt, red ribbon, smoke haze, natural lighting, expressive, digital art, casual setting
Example 5：
1girl, solo, chestnut hair, single braid, white blouse, black jacket, orange grove, smiling, happy expression, sunshine, dappled light, orchard, nature, cheerful, warm light, painterly style, outdoors, lush greenery, casual outfit
Example 6：
1girl, close-up, blue hair, long hair, white rabbit, red eyes, blood on face, terrified expression, horror, fear, trembling, gasping, cinematic lighting, pale skin, dramatic shadows, emotional, digital art, close-up portrait, intense atmosphere, expressive style
Example 7：
1girl, school uniform, black long hair, straight hair, crossed arms, smug expression, bullying vibe, dominant pose, low angle, stairwell, school setting, superiority, cold lighting, psychological tension, digital art, dramatic composition
Example 8：
The dreamy cartoon illustration depicts a blonde girl lying in a field of gorgeous flowers with a smile, basking in the warm sunshine. The foreground highlights her happy face and upper body, with her eyes closed and looking up at the surrounding pink, purple, red and blue flowers.She wore a soft beige dress with a floral brooch in the neckline. The sun filtered through the petals, casting a mottled glow that created a soft, glowing atmosphere.The background is rich in blurred flowers and green, which enhances the sense of peace and happiness.The painting style is bright, the color tone is soft, the gradient is delicate, and the overall atmosphere is warm and peaceful, full of spring beauty.
5.5 Limb movement control
• Try to think about the orientation and movement of your head and limbs before you write.
• Large body movements often involve changes in perspective to enhance the tension of the movement and reflect the rationality of the character's movement.
Example 1：
dynamic_perspective, dynamic_pose, ink_wash_eastern_dragon, eastern_dragon_surrounds_sword, EN Description: A boy leaps mid-air in a dynamic ink-wash painting style, her glowing skin illuminated by backlight that outlines disheveled hair whipping in the wind. The monochrome composition erupts with a crimson blade at its core, surrounded by swirling Eastern dragon ink textures that dissolve into colorful splashes. Stark chiaroscuro contrasts her fierce fighting stance – captured in dramatic foreshortening – with the dark heme background, cinematic lighting carving her focused expression as if frozen in a frame from an animated wuxia epic. Neon accents shimmer along the sword's edge where the dragon's form merges with the weapon, traditional brushstrokes blending seamlessly with cyber-ninja aesthetics.
Example 2：
pillow_hug, standing, half-closed_eyes, rubbing_eye, one_eye_closed, sleepy_expression, bedhead_hair, oversized_sweater, loose_pajamas, soft_lighting, morning_ambiance, drowsy_pose, cheek_on_pillow,yawning, stretching_pose, warm_shadows, sunbeam_highlight A girl stands drowsily hugging a plush pillow, one eye squeezed shut while rubbing the other with a knuckle. Her half-lidded gaze blurs with sleepiness, bed-tousled hair falling over an oversized sweater as morning light filters through nearby curtains. The pillow nearly swallows her frame, its fabric creased where her cheek rests in a moment of tender vulnerability between sleep and wakefulness.
Example 3：
A stunning and energetic digital painting in a thick paint (厚涂) style, capturing the signature radiant lighting and dynamic feel of the artist Alphonse (white datura). The artwork features a lone vampire ballerina, her body a study in graceful power and supernatural elegance.
She is captured mid-performance, executing an incredibly high leg lift or kick. Her barefoot leg is extended dynamically, showcasing the arch of her foot and her pointed toes. Her very long, floating blonde hair swirls around her, its movement echoing the energy of her dance. She looks directly at the viewer with a captivating smile, her mouth slightly open to reveal subtle, sharp fangs. Her luminous golden-orange eyes glow with a mischievous light, and her pointy ears peek through her elegant sidelocks.
She wears a simple yet beautiful white ballet dress, possibly with a skirt cutout that enhances her long lines. A key visual element is the flurry of pink ribbons, which wrap around her limbs and float weightlessly through the air, their paths weaving around her and adding to the composition's incredible sense of motion. The scene is illuminated by dramatic, warm, directional lighting that bathes her skin in a soft, luminous glow, beautifully highlighting her collarbone and the lean muscles of her form against a dark,atmospheric background.
极致的细节, 精彩的光影, 动态光影,
超高对比度，明暗对比
厚涂3d质感,极致的细节,精彩的光影, 动态光影,极致的景深效果,高对比度,色彩丰富,精细光线,漏光效果,焦外虚化,前景模糊,光线追踪,镜头光晕,电影级打光，超写实背景,光晕，线条清晰,明暗对比,超高清,最高画质,强光,细腻渲染,丰富的细节,高质量,写实逼真,超清晰,
masterpiece, best quality,
Example 4：
This is a dynamic digital illustration filled with strength and beauty. At the center of the frame, a young woman with a well-toned physique sits gracefully on the floor in a perfect middle split, showcasing incredible flexibility and power. Her body is perfectly aligned with her torso upright, her arms raised and hands pressed together above her head in a prayer position, fingers gently touching. Her expression is focused and serene, exuding grace and self-control.She is dressed in a casual outfit: a simple white t-shirt and high-waisted denim blue shorts that outline her fit physique. Soft, natural light bathes the scene, highlighting the graceful curves of her toned arms, legs, and waist while casting gentle shadows. The background is simple, featuring a clean wooden floor, which emphasizes the stability and balance of her grounded pose. The smooth digital art style focuses on her elegant form, her posture, and the calm, tranquil mood of the scene, conveying a unique charm that blends strength, flexibility, and elegance.
Example 5：
"Graceful digital illustration of a ballerina practicing in a studio, captured mid-pose in a classic ballet position, en pointe. The ballerina has long, flowing brown hair, tied back in a neat, elegant bun. She wears a delicate, form-fitting white tutu with layers of soft tulle, the skirt fluttering slightly as she moves. Her tight ballet bodice is a pale pink, with subtle lace details along the neckline and shoulders, accentuating her graceful silhouette. She stands on her toes in an elegant arabesque, with one leg extended horizontally behind her, held in perfect form, while the other leg is bent at the knee and raised slightly to the side. Her arms are poised beautifully, one arm raised overhead in a gentle arc, the other extended in front of her at shoulder height, fingers delicately stretched to convey an ethereal, flowing quality. Her back is straight, chest lifted, and her gaze is focused ahead, exuding determination and grace. The ballet studio is softly lit, with wooden floors and mirrors lining the walls, reflecting her poised figure. The lighting emphasizes the soft folds of her tutu and the fluidity of her movement, creating an atmosphere of serenity and elegance. The digital art style is refined and painterly, with a focus on the soft textures of her attire, the fluidity of her pose, and the overall graceful ambiance of the scene."
Example 6：
"Dynamic sports anime illustration of a young male volleyball player in mid-action diving for a ball. Character has spiky dark hair with blonde forelock and determined yellow/gold eyes. Player wears orange and black volleyball uniform with number visible on jersey, knee pads, and athletic shoes. Full body shot capturing athletic form in horizontal diving pose with arms outstretched. Minimal white background emphasizing the character's movement with subtle motion lines. Volleyball visible at edge of composition. Art style features loose, energetic brushwork with minimal coloring focusing on oranges and blacks. Character shows concentrated, intense facial expression with sweat details. Inspired by Haikyuu!! anime style character design and dynamic sports action poses. Sketch-like quality with emphasis on capturing the explosive movement and athleticism of volleyball. Warm color palette dominated by orange tones creating visual energy and excitement."
Example 7：
The upward view can reflect the sense of authority, and the position and character of the person can be reflected to a certain extent with the action of crossing legs.
"Dramatic digital illustration of a long-haired, handsome mafia boss seated with legs crossed, viewed from a low-angle upward perspective. He exudes authority and danger, wearing a sharp, dark tailored suit with a subtle pinstripe pattern, the jacket slightly open to reveal a dark vest and crisp collared shirt beneath. His long, flowing black hair cascades over his shoulders, some strands catching the ambient light. In one hand, he casually holds a lit cigar between his fingers, with a faint trail of smoke curling upward. His gaze looks down from above, brows slightly furrowed, lips set in a firm line, conveying displeasure or irritation. The low-angle viewpoint emphasizes his dominance, power, and intimidating presence. His other arm rests confidently on the armrest or his knee. The background is dim and atmospheric, possibly the interior of a lavish private lounge or office, with hints of gold trim, leather furniture, and shadowy lighting. The digital art style is refined and cinematic, using sharp contrasts, detailed facial expression, and dramatic lighting to heighten tension and elegance."
Example 8：
"Minimalist anime illustration with white background showing a young man with black hair in an unusual contorted position. Full body view of character wearing an oversized gray/beige trench coat or overcoat with belt detail, black pants, and formal black shoes. Subject is bent forward with hands touching the ground in a creepy, unnatural pose suggesting supernatural or horror themes. Character has pale skin and dark hair partially covering face, creating an ominous appearance. Clean art style with simple shading and minimal details focusing on the disturbing body position. Monochromatic color scheme dominated by grays and blacks against stark white background. No environmental elements or context, emphasizing the isolated, eerie quality of the figure. Japanese horror anime aesthetic with psychological thriller undertones. Professional digital illustration with precise line work and controlled color palette. Character design suggests themes of possession, psychological distress, or supernatural abilities."
VI: Quality optimization and common problem resolving
6.1 Quality improvement strategy
Current valid quality terms:
• masterpiece,best quality-basic quality assurance
Recommended quality word combinations:
masterpiece,best quality,6.2 Common problems and solutions
Problem 1: Character traits are not obvious
• Solution: Enhance the tag and natural language parts of the character description; use # to strengthen the character.
• Optimization: Replace simple labels with specific physical descriptions
Problem 2: Background and role are not in harmony
• Solution: Clearly describe the relationship between the character and the environment in natural language
• Optimization: stand confidently in the center of the magical garden
Problem 3: inconsistent style
• Solution: Limit the number of artist tags, strengthen the style description; use @ to enhance the artist's style.
• Optimization: Select 1-2 main artists and describe the unified style in natural language
6.3 Negative cueing strategy
Compared to SDXL, Lumina2 is more sensitive to negative prompts, which means they are more effective in removing unnecessary elements, but if used improperly, they are also more likely to suppress the original character characteristics and art style of the model.
Therefore, it is recommended to start with the following prompts that have been tested and have less impact on the picture style. The following combination can effectively optimize the picture in most cases while preserving the characteristics of the original character to the greatest extent.
Basic negative word combinations:
You are an assistant designed to generate anime images based on textual prompts. <Prompt Start> blurry, worst quality, low quality, deformed hands, bad anatomy, extra limbs, poorly drawn face, mutated, extra eyes, bad proportionsAdvanced negative word combinations:
You are an assistant designed to generate images based on textual prompts. <Prompt Start>
A low quality, low resolution, ugly, and disgusting image with severe digital artifacts, blur, and noise. The subject is deformed, disfigured, and malformed, with bad anatomy, mutated limbs, and bad proportions. Features distorted, twisted, and unnatural hands and face, with extra or missing fingers. The style is amateurish, poorly drawn, childish, like a flat, unfinished sketch or a cheap, bad CGI render. The composition is bad, cropped, out of frame, and cluttered with text, watermarks, or signatures,