AI 提示词库

young beautiful woman scientist, 22-25 years old, curly long brown hair, freckles, striking blue eyes, wearing white lab coat over black top, delicate silver necklace, standing in the most advanced futuristic AI laboratory, high-tech glowing servers, holographic displays, blue and neon lights, quantum computers, glass walls, robotic arms, waiting patiently with calm confident expression, ultra-realistic, cinematic lighting, highly detailed, 8k

# Role: Comparison Infographic Generator (Design-Centric with Toggles) # === USER INPUT SETTINGS (FILL THESE IN) === # 1. Grid Size (e.g., "2x2 (4 panels)", "3x3 (9 panels)"): {argument name="grid size" default="1x5"} # 2. Base Subject Description (Identical base scene for all panels): [{argument name="subject description" default="Cold white, slender, fair Asian female legs, skin delicate and smooth, without blemishes, almost invisible pores and hair follicles. Skin is not dry, slightly moist"}] # 3. Variations Theme (What is being compared?): {argument name="variation theme" default="Comparison of bare legs with different stocking materials and styles"} The feet need to be fully displayed in the frame # 4. Show Text Overlay at Bottom (Yes/No): NO # =========================================== # Task Description: Generate a multi-panel infographic grid chart for a comparative visual study based on the User Input Settings above. The goal is to showcase different textural or stylistic variations (defined in the data list below) applied to the identical Base Subject defined above. # Global Constraints & Style: 1. **Layout:** A **[Insert setting from "1. Grid Size" here]** infographic grid chart. 2. **Consistency is Key:** The Base Subject, camera angle, composition, and lighting must remain absolutely identical across ALL panels to emphasize only the specific variations defined in the data list. 3. **Photography Style:** Clean, controlled studio photography with neutral, even lighting designed to highlight texture differences. # Text Overlay Display Logic (CONDITIONAL): (AI Instructions: First, check the value of User Input Setting "# 4. Show Text Overlay at Bottom". You MUST execute ONLY ONE of the following logic blocks based on that value.) ##### ➤ LOGIC PATH A: IF Setting #4 is "Yes" (Show Overlay) If the user selected "Yes", EVERY panel must feature a precisely designed text overlay at the bottom following these strict rules: * **Background:** A clean, sharply defined solid rectangular color block (e.g., charcoal grey or black) spanning the entire width of the panel's bottom section. * **Typography:** Inside this block, the text must be rendered in an **elegant, ultra-thin, minimalist sans-serif font**. * **Layout:** The text "[Number]. [Short Title] (Prompt: [Key English Prompt Snippet])" must be perfectly centered within the color block. * **Aesthetic:** The overall effect should be that of professional, minimalist graphic design, with crisp lines and perfect alignment. ##### ➤ LOGIC PATH B: IF Setting #4 is "No" (Hide Overlay) If the user selected "No", apply this CRITICAL rule: * **CRITICAL:** DO NOT add any text overlay, numbers, titles, background color blocks, or prompt snippets to the images. * **Result:** The bottom of every panel must be clean image data only, showing the subject fully without obstruction. # Base Subject Definition: The base scene for every panel is: **[Insert description from "2. Base Subject Description" here]**. The variations below will modify the surface or appearance of this base subject. # Variations Data List (Apply sequentially to the grid): (AI instructions: Iterate through this list based on the required grid size. For each panel, apply the user-defined 'Visual Change' to the base

Perfect cinematic cross-section cut through {argument name="dish" default="[DISH]"}, sliced cleanly down the center revealing every internal layer in full detail. The cut face is facing the camera straight-on, like a geological diagram of flavor. Each visible layer is labeled with a thin horizontal line extending to the right — uppercase sans-serif "{argument name="ingredient name" default="[INGREDIENT NAME]"}" with italic descriptor below. Background: flat matte {argument name="color" default="[COLOR]"}. Lighting: dramatic side light from left casting long shadows, single warm spotlight on the cut face. Every texture hyperreal — visible moisture, air pockets, char, glaze, and steam. Mood: architectural food photography meets scientific illustration. Dish name "[DISH NAME]" in large elegant type at top. Subtitle: "[TAGLINE]". 4K, tack sharp, Style ref: Wes Anderson symmetry meets Michelin-star kitchen.

[Nano Banana Pro Compatible] AI Utilization Techniques to Turn Anything into a Diagram: Convert to a 'Diagram that Conveys in 2 Seconds'! Complete Guide to Image Generation Prompts for Business

Please help me understand this paper. Mark important content (figures or formulas) with a circle and add annotations so that I can understand it accurately (in a hand-drawn style).

{ "image_description": { "overall_theme": "High-technology augmented reality analysis of a scenic outdoor portrait", "resolution": "4K UHD, hyper-realistic with sharp data overlay", "subject": { "name": "Woman_Model_A", "pose": "Seated_profile, looking_left, side", "attire": { "garment": "Draped_One_Shoulder Pleated_Silk_Green_Dress", "color_hex": "#1E6641", "footwear": "Boots_Maroon", "accessory": { "type": "Bag_Two-Tone Leather/Metal", "details": "Gold_Buckle_feature" } } }, "environment": { "location": "Nusa_Penida_Cliffside", "terrain": "Rocky_Path", "details": { "main_trees": { "type": "Fern_Tree", "count": 3 }, "vegetation": "Lush mossy trees, vibrant red flowers, exotic plants", "boundary": "Wooden fence posts with thick rope rail", "background": "Distant tree-covered cliffs partially obscured by mist, with birds in flight", "specific_sign": "Small sign on a fence post with the text 'Kelingking'" } }, "data_overlay": { "style": "Translucent cyan digital interface with glowing gridlines and panel outlines", "panels": [ { "position": "Top-Left (Stacked Vertical)", "content": [ "ANALYSIS: SUBJECT_DETECT", "POSE: SEATED_PROFILE", "OUTFIT: DRAPED_ONE_SHOULDER", "COLOR: #1E6641_SILK_GREEN", "ACCESSORY_1: BAG_TWO-TONE", "MATERIAL: LEATHER/METAL", "HEEL: BOOTS_MAROON" ] }, { "position": "Top-Center (Floating Block)", "content": "COORDINATES: -8.7495° S, 115.0934° E" }, { "position": "Center-Left (Embedded Annotation)", "linked_to": "Leftmost Fern Tree", "content": "FERN_TREE (count: 3)" }, { "position": "Center-Right (Embedded Annotation)", "linked_to": "Rightmost Fern Tree", "content": "FERN_TREE (count: 3)" }, { "position": "Top-Right (Annotation)", "linked_to": "Cliffs", "content": "CLIFF_DISTANCE (approx: 450m)" }, { "position": "Bottom-Right (Formatted JSON Block)", "content": "{\n \"model\": {\n \"name\": \"Woman_Model_A\",\n \"pose\": \"Seated_left_looking_side\"\n },\n \"apparel\": {\n \"garment\": \"Pleated_Silk_Green_Dress\",\n \"style\": \"One-Shoulder_Draped\"\n },\n \"bag\": {\n \"type\": \"Two-Tone_Leather_Gold-Buckle\",\n \"brand\": \"Luxury_Accessory\"\n },\n \"location\": \"Nusa_Penida_Cliffside\",\n \"terrain\": \"Rocky_Path\"\n}" } ] } } }

<instruction> Inputs: Ingredients = [{argument name="ingredient 1" default="Chicken"}, {argument name="ingredient 2" default="Bacon"}] Step 1: Global Analysis Analyze the ingredients. Select 4 Distinct Countries with a famous dish using these ingredients Goal: A "Culinary Atlas" Grid showing 4 different cultural variations of the ingredients. Layout Rule: Output Format: A 2x2 Grid. Content per Panel: Each of the 4 panels must show a Single Open Book Spread for that specific country. Inside Each Panel (The Book Spread Rule): Left Page (The History): 2D Vintage Sepia Sketches. A visual timeline or illustrated recipe of the dish (e.g., "1850 -> 1920 -> Today"). Flat, ink-drawn style. Right Page (The Reality): 3D Pop-up Diorama. The finished, steaming meal emerges physically from the page. A tiny miniature chef from that country is standing on the page next to the massive food. Consistency: All 4 books should look like part of the same series, but with different cultural contents. Lighting: Cinematic top-down lighting, highlighting the depth difference between the flat left page and the 3D right page. Output: 2x2 Grid, High Definition, Macro Detail. </instruction>

Referencing this character, generate a Pokémon Zukan-style illustration. The Zukan No. is {argument name="Zukan Number" default="[Enter Zukan No.]"}. The background is white. The Zukan description text is {argument name="Description Text" default="[Enter description text]"}. Reproduce the Zukan layout and design.

一幅黑白线条画，中间以醒目的粗体泡泡字体呈现文本“{argument name="quote" default="DOING MY BEST RESULTS MAY VARY"}”。文本周围环绕着各种可爱且奇趣的元素，包括微笑的星星、猫咪脸、云朵、雏菊、爱心、纸杯蛋糕、波霸奶茶以及闪烁的星芒。整体构图紧凑且充满童趣，专为涂色页设计。

[Character Reference Specification] Must use the character from the reference image. Must depict the same person in all panels. Facial features, hairstyle, eye shape, and atmosphere must be perfectly consistent. Changing the person, face, or hairstyle is prohibited. Clothing may change for each day of the week, but the character itself must be the same. Even if deformed, the character must be instantly recognizable as the person in the reference image. [Overall Concept] An anime illustration representing one week from Monday to Sunday, in a perfectly circular shape. [Composition/Layout] ・The entire circle must be divided into exactly 7 pizza-cut segments. ・6 or 12 divisions are prohibited. ・Exactly 7 equal sectors. ・Infographic-style layout. ・All 7 panels must connect naturally. [Order of Days (Strictly Enforced - Clockwise)] MONDAY → TUESDAY → WEDNESDAY → THURSDAY → FRIDAY → SATURDAY → SUNDAY → returns to MONDAY [Circulation Expression] ・Place a thick circular arrow on the outer circumference of the circle. ・Clockwise direction. ・Emphasize that the week repeats infinitely. [Character Design] ・Comical anime character with strong deformation. ・Chibi-style body shape. ・Large head and small body. ・Short limbs. ・Expressions are extreme and gag-oriented. ・Cute and exaggerated reactions. [Text Rules] ・Only English day names are placed in each panel: "MONDAY", "TUESDAY", "WEDNESDAY", "THURSDAY", "FRIDAY", "SATURDAY", "SUNDAY". ・No dialogue, speech bubbles, or Japanese text whatsoever. ・Day names must be bold, easy to read, and without distortion. ・Placed along the shape of each panel. [Depiction of Each Day] MONDAY: Deformed office worker (OL) in a suit. Messy hair, dead eyes. Holding two large coffees. TUESDAY: Working silently on a computer. Surrounded by a mountain of documents taller than herself. WEDNESDAY: Working late overtime. Dark office, night view outside the window. Slumped posture, utterly exhausted. THURSDAY: Comical expression of the soul leaving the mouth. A transparent, cute soul floats up from the body. Depicting extreme fatigue as a gag. FRIDAY: High-tension expression. Taking off her jacket, Holding a large beer mug overflowing with foam. Festive atmosphere. SATURDAY: Out happily with friends in casual clothes. Energetic pose, jumping. Bright and lively atmosphere. SUNDAY: Lying down on the bed as if collapsing. The word "MONDAY" is visible on the smartphone screen. Exaggerated despair. [Art Style/Quality] ・Comical anime style with strong deformation. ・Thick line art. ・Simple shading. ・Bright and pop color scheme. ・Expressions are very easy to understand. ・High-quality illustration. [Negative Prompts] ・Realistic style, photorealistic expression. ・Changes in character's face or hairstyle. ・Composition other than 7 divisions. ・Disruption of the day order. ・Japanese day names. ・Dialogue, speech bubbles. ・Distortion or collapse of text. ・Square panel division.

Create an info graphic about the 1969 Woodstock Festival

完全重塑上传的人物。 将其作为一篇笔记文章的标题图片，该人物在文章中介绍“Nano Banana Pro”。 长宽比：横向 16:9。 风格和颜色：简洁、手绘风格、斜体，带有蓝绿色渐变。 标题文字：“深入解读 Google 全新 AI ‘Nano Banana Pro’”。

[NATURAL OBJECT] dissected as if by a master naturalist who found it in the wild. Split and pinned open like a specimen, but lit as if Caravaggio worked for National Geographic. Every internal structure glowing with its own material truth. Cross-section so sharp it feels violent. Interior so beautiful it feels sacred. The full object present, one half intact, one half opened to its core, every layer at the resolution of obsession. Background: pure black velvet. The object floats in it like something precious and dangerous. Labels pressed close, handwritten-weight serif, never floating: "[STRUCTURE]" / "[MATERIAL TRUTH]" / "[AGE OR SCALE]" "[OBJECT] DISSECTED" title in warm ivory caps, upper left. "[One line that makes you lean closer]" italic beneath. Mood: Audubon illustration meets Caravaggio meets the most beautiful science photograph ever taken. 4K, specimen lighting, extraordinary interior detail, nothing clinical, everything alive.

A {argument name="object_en" default="metal ball"} dropping into water, with a slow-motion splash, realistic physics and refraction, at {argument name="resolution_en" default="2K"} resolution.

Do this for {argument name="scientist 1" default="Einstein"} and a competitor of his <instructions> Invention = [{argument name="invention" default="Electricity"}] Phase 1: Historical Conflict Analysis 1. Identify Rivals: Determine the two main historical figures associated with this invention (e.g., Tesla vs. Edison, Wright Bros vs. Curtiss). 2. Identify The Tech: Distinguish their competing methods (e.g., AC Coils vs. DC Bulbs). 3. Identify The Era: Select appropriate "Giant" background clutter (e.g., Victorian newspapers, 1920s fountain pens). Phase 2: Visual Execution Goal: A "Macro Diorama" of a Rivalry. A messy, highly detailed flat-lay of a shared historical workspace where the two rivals exist as tiny living figurines battling over their creation. Rules: The Stage: A large, yellowed Patent Diagram or Blueprint lies flat on a wooden desk. The paper is wrinkled and covered in scribbled corrections. The Tiny Rivals: Two hyper-realistic miniature figurines (Inventor A and Rival B) are standing on opposite sides of the document. They are positioned at their own tiny wooden desks or workstations on top of the paper. The Conflict: They are gesturing aggressively at each other. The Invention itself is physically emerging from the center of the paper between them (e.g., a glowing 3D lightbulb or a spinning biplane propeller), acting as the barrier or prize they are fighting over. The "Giant" Terrain (Surroundings): The scene is surrounded by massive, period-accurate objects acting as a landscape: A "Giant" wall of newspapers with headlines about their legal battles. Massive burnt fuses or broken prototypes scattered like boulders. A spilled cup of coffee creating a "lake" near one of the figures. Giant tools (calipers, screwdrivers) casting long shadows. Perspective: Slightly angled 45-degree macro shot looking down at the patent. Texture: Dust motes, scratched mahogany wood, paper grain, metallic reflections. Lighting: Cinematic Macro Lighting. A warm desk lamp illuminates the scene, but the Invention in the center emits its own light (e.g., blue electrical arcs vs. yellow filament glow), lighting the faces of the tiny figures. Output: ONE image, 4:5 aspect ratio, photorealistic historical macro photography, tilt-shift effect. </instructions>

分析产品并创建一个高分辨率、照片级真实感的专业广告参考表，包含以下四个部分，背景为干净的白色/中性摄影棚背景，采用高端商业产品摄影风格，柔和可控的灯光，逼真的材质，清晰的细节，微妙的阴影以及专业的编辑网格布局。 1. 产品概览；名称、类别、材质、尺寸/大小、主要功能和品牌颜色。 2. 产品用途/主视图 - 正面、背面、侧面、3/4 视角，以及相关的功能/使用视图。 3. 产品特写 - 标志/标签、纹理/材质、关键组件、包装细节以及其他重要的识别特征。 4. 包装 / 结构 - 适用的相关包装或结构视图。 整个参考表必须作为 AI 图像/视频制作的视觉连续性指南。每个产品视图都必须以绝对一致的方式描绘完全相同的物理产品。不要重新设计、添加变体、更改品牌标识或改变比例。使用简洁的排版、清晰的部分标签、细分隔线、平衡的间距和精致的制作设计美学。

Imagine a pencil tip seen through a microscope. Imagine how the pencil tip would look and create an image.

<instruction> Input A is a world-changing invention (photo or invention name). Analyze the inventor, eureka moment, prototype design, and societal impact. Optional Input B is a style reference (patent office / inventor's workshop). If missing, choose steampunk-meets-modern museum aesthetic. Goal: Wooden display box with brass hinges containing the miniature "first prototype" surrounded by inventor's tools, failed attempts, and blueprint fragments. Rules: - Center: tiny working-looking prototype of the invention - Include inventor figurine at miniature workbench - Surround with 10-15 period tools, sketches, failed versions - Add aged paper "patent document" visible through glass top - Materials: dark wood, brass accents, vintage glass - Plaque with invention name, inventor, patent year Output: ONE image, 1:1 inventor's journal aesthetic </instruction>

Motion Prompt: The sandwich start to spin while ingredients separate gently and precisely, maintaining alignment and scale. The motion is smooth, with no extra effects. IMAGE:— Using the first breakfast sandwich image as a base, create a clean exploded-view visualization of the same sandwich. 5 elements: – Bread slices – Egg – Bacon – Cheese – Sauce Annotations: – Bread: “Toasted structure” – Egg: “Soft protein layer” – Bacon: “Crispy smoky flavor” – Cheese: “Melty richness” – Sauce: “Savory moisture boost”

Ancient Egyptian profile portrait of the person in the reference image, side-facing pose (strict profile view), inspired by {argument name="style source" default="traditional tomb paintings"}, {argument name="features" default="almond-shaped eye, elongated eyeliner, sharp nose and chin"}, wearing a {argument name="headwear" default="nemes headdress"} and ornate gold jewelry, flat perspective, stylized proportions, warm sandstone and gold color palette, hieroglyphic background, smooth matte texture, soft even lighting, highly detailed, clean lines, 4k, museum-quality artwork

{ "title": "MY 2025 SCRAPBOOK", "theme": "Personal Year in Review / Mosaic Stone Aesthetic", "visual_style": "Textured stone background with inlaid illustrations representing different memories.", "memories": [ { "category": "Biggest Achievement", "content": "Ran a marathon", "visual_description": "A man raising his hands in victory on a ribbon banner." }, { "category": "Overcame", "content": "Procrastination", "visual_description": "A man breaking free from chains." }, { "category": "Unlocked", "content": "Gymnastic skills", "visual_description": "A man performing a handstand on a green book." }, { "category": "Core Memory", "content": "Grandma's 75th Birthday!", "visual_description": "A framed picture of the man hugging an elderly woman with a birthday cake." }, { "category": "One for the soul", "content": "First Solo trip", "visual_description": "A hiker with a backpack looking toward a mountain peak." }, { "category": "Trip Memory", "content": "Bali", "visual_description": "A ticket stub illustration showing the man near a traditional Balinese temple." }, { "category": "I ate the most", "content": "Ramen", "visual_description": "A man eating noodles with chopsticks from a bowl." }, { "category": "Overused Gemini prompt", "content": "Plan my schedule", "visual_description": "A man working on a laptop." }, { "category": "My motto for 2026", "content": "Visit Italy", "visual_description": "A toy rocket ship blasting off." } ] }

Create high-fashion couture sketches for an "{argument name="collection name" default="Aria of Frost"}" and "Lumina Papilio" collection, featuring ethereal, butterfly-inspired evening gowns. The designs should showcase sleek mermaid silhouettes, architectural ruffled sleeves that mimic lepidoptera wings, and intricate silver filigree wirework. Use a luxurious color palette of {argument name="color palette" default="iridescent amethyst, frozen lilac, and deep cerulean blue"} accented with metallic gold and silver threads. The mood board should include technical details like fabric swatches of silk-satin, silver mesh, and seed pearl clusters, presented in a professional atelier style with elegant calligraphy and butterfly reference anatomy.

HUMAN BODY ultra-clean anatomy infographic

创作一张 4:5 竖构图的高级复古日本旅行日志编辑拼贴画，包含四个不同的城市板块——{argument name="locations" default="泰国曼谷；葡萄牙里斯本；摩洛哥马拉喀什；土耳其伊斯坦布尔"}。每个板块将温暖的照片级真实感纪实街头摄影与相匹配的、带有瑕疵的手绘丝网印刷插图相结合，非对称地排列在做旧奶油色纸张上，并配有撕纸层、档案边框、细微的铅笔痕迹、打字机标题、褪色的墨水、纸张纤维以及真实的 35mm 胶片颗粒感。曼谷：神龛、缠绕的电线、热带绿植、青色摩托车；里斯本：黄色电车、陡峭的鹅卵石街道、阳台和河流景观；马拉喀什：赤陶色麦地那、棕榈树、地毯和华丽的宣礼塔；伊斯坦布尔：陡峭的石街、加拉塔塔、博斯普鲁斯海峡和复古踏板车。使用鼠尾草绿、赭石色、赤陶色、芥末黄、炭灰色和暖象牙色的克制色调，呈现精致的编辑艺术指导、柔和的对比度、自然光、手工制作的瑕疵以及博物馆级的旅行杂志美学。小型档案标签：“Bangkok / Field Note 052”、“Lisbon / Field Note 031”、“Marrakech / Field Note 017”、“Istanbul / Field Note 039”。无 Logo、水印、HDR、过度饱和或现代杂乱元素。