Galería de prompts

A sophisticated commercial photograph, set against the backdrop of a modern transportation hub or public waiting area, is shot from a frontal, slightly low angle. On the left and in the center of the frame is a giant, illuminated Pepsi vending machine. A young man in a light gray hoodie, black athletic shorts, white mid-calf socks, and white sneakers, with his back to the camera, stretches upwards, attempting to grab a bottle of Pepsi from the top shelf. He stands precariously on four red Coca-Cola cans stacked vertically on the floor, a visually humorous gesture conveying the message that "it's worth the extra steps for Pepsi." The vending machine has ample shelves, with bright, cool white lighting inside: the top shelf displays multiple bottles of Pepsi, the middle shelf showcases various clear, green, yellow, and amber bottled beverages, and the bottom shelf holds rows of red Coca-Cola cans. The vending machine is Pepsi blue, with a large Pepsi globe logo and the word "PEPSI" printed on the left panel, and a payment/control panel on the right. At the top of the image is a large, bold, white sans-serif headline.{argument name="headline text" default="WORTH THE EXTRA STEPS."} Center it on the dark-colored upper wall. Add a simple white sans-serif brand tagline at the bottom center.{argument name="tagline text" default="Choose Pepsi."} On the right wall hangs a framed Pepsi poster with the bold, stacked text "REFRESH RECHARGE REPEAT" and the Pepsi logo printed above it. A simple metal bench sits below the poster. The image employs a bright, reflective surface, sharp focus, a high-quality advertising photography feel, realistic proportions, high dynamic range, and a brand-safe composition that creates a strong contrast between Pepsi blue, Coca-Cola red, and neutral gray.

A faceted rectangular glass perfume bottle filled with pale amber liquid, standing on a dark matte surface against a pure black background. Three crisp cubic shafts of warm white light cut diagonally across the bottle, producing hard-edge reflections on the glass facets and sharp highlights on the gold metal cap. No label, no brand text, no watermark. Square 1:1 hero composition, luxury product photography, photorealistic glass refractions, ultra detailed.

基于此角色和背景，请制作一份类似官方设定资料的角色资料卡。 ・包含三视图：正面、侧面和背面 ・添加角色面部表情的变化・分解并展示服装和装备的详细部分 ・添加色板・包含世界观设定的简要说明 ・总体上，使用有组织的布局（白色背景，插画风格）高分辨率、专业概念艺术风格

Create a subject{argument name="reference" default="my look"} Inspired Minecraft Skins

Goal: Create an ornate 3D pop-up storybook scene where {argument name="character name" default="Nozomu"} appears to leap out of an open fairy-tale book, as if a custom character illustration has become a luxurious paper craft diorama. Canvas: Square composition, warm cinematic close-up view from slightly above, centered on an open book resting on a dark antique wooden desk. Use shallow depth of field, glowing candlelight, amber highlights, soft shadows, and a magical vintage atmosphere. Main subject: A beautiful anime-style young woman in an extravagant {argument name="dress color" default="deep red"} ball gown sits/steps forward from the pop-up pages. She has long softly curled brown hair, delicate jewelry, bare shoulders, lace gloves, and layered translucent ruffles with gold sparkles. Her face should be intentionally covered by a soft square blur/mosaic privacy mask while the rest of the illustration remains sharp and detailed. Pop-up book structure: Build the scene from layered cut-paper pieces with thick white paper edges, scalloped borders, tabs, folds, and dimensional shadows. Include exactly 8 major pop-up elements: 1 large central character cutout, 1 curved title sign at the upper left, 1 moon-and-stars night-sky arch behind the character, 1 cream vintage sofa on the right, 1 tall arched window on the far right, 1 small purple castle on the left, 1 ornate chapter placard on the lower right, and 1 central folded stair/path bridge rising from the open pages toward the character. Decorative details: Use abundant roses, vines, tiny stars, clouds, gold filigree, ribbons, lace-like paper trims, and scattered rose petals. Surround the book with antique props including candles, a clock, dark floral decor, and gilded frames, all blurred slightly in the background. Text content: Use elegant Japanese storybook typography printed directly on the paper. The upper-left title sign should read 「ようこそ、のぞむの世界へ」 with the subtitle 「夢と魔法のはじまり」. Add exactly 4 readable story text areas: 1 small name plaque on the left reading 「のぞむ」, 1 large left-page story panel, 1 right-page story panel, and 1 lower-right chapter card reading 「第1章 のぞむと赤いドレスの夜」. Add a bottom ribbon label reading 「のぞむのものがたり」. Keep the Japanese text decorative and believable, not overly crisp. Style: Highly detailed anime illustration mixed with realistic handcrafted pop-up book photography, romantic princess fantasy, rococo ornamentation, warm beige parchment paper, deep reds, gold accents, dusky purple night sky, magical sparkles, premium children’s picture-book aesthetic. Constraints: The open book must be the main object, not a flat poster. Maintain strong 3D paper depth and visible layered construction. Do not add extra characters. Do not include watermarks, logos, modern objects, or English text.

Based on our previous conversation, create a design that suits my personal style.{argument name="character type" default="VTuber"} image.

{ "type": "Brand Visual Identity System Project", "brand": { "name": "{argument name=\"brand name\" default=\"dog\"}", "english_mark": "{argument name=\"logo letters\" default=\"GDX\"}", "industry": "{argument name=\"industry\" default=\"pet industry\"}", "date": "{argument name=\"date\" default=\"2024.05\"}", "tagline": "Love it, understand it, accompany it." }, "style": { "overall": "The minimalist, high-end corporate display style features a clean Swiss-inspired layout, ample white space, light gray dividing lines, and a soft, neutral background.", "palette": "Deep forest green and warm off-white, accented with a few accent color blocks.", "mood": "Modern, trustworthy, pet-friendly, sophisticated" }, "canvas": { "aspect_ratio": "3:4 Vertical Version", "background": "Soft warm white" }, "header": { "left_title_cn": "Brand visual identity system", "left_title_en": "BRAND IDENTITY SYSTEM", "right_tagline": "Love it, understand it, accompany it.", "center_logo": "A large, custom-designed GDX lettering, featuring a white dog silhouette in the center, written in bold geometric curves and diagonal strokes, in a dark green color.", "center_subtitle": "dog" }, "layout": { "sections": [ { "title": "Basic Information", "position": "Top left", "count": 3, "labels": [ "Brand Name", "Industry attributes", "Design time" ] }, { "title": "Design grid", "position": "Middle left top", "count": 1, "labels": [ "A logo with a scale of 1.618:1 is constructed using a grid." ] }, { "title": "Concept sketch", "position": "Upper-middle", "count": 4, "labels": [ "Dog head and letter sketch", "Contour optimization", "Intermediate combination marker", "Final simplified tag" ] }, { "title": "Inspiration", "position": "Middle right top", "count": 6, "labels": [ "Minimalist white architectural arch", "Golden Retriever side profile", "Curved arch combination", "Close-up of green leaves", "Dark green material cube", "Light-colored wood grain squares" ] }, { "title": "Creative concept", "position": "center left", "count": 4, "labels": [ "Design Philosophy and Symbolic Meaning", "Brand positioning services", "Color Psychology and Strategies", "Scalability and Adaptability" ] }, { "title": "Brand Application", "position": "From center to right", "count": 7, "labels": [ "front and back of the business card", "letter paper and envelope", "APP icon", "Website Homepage / Website Icon", "Product packaging / shopping bag", "Storefront sign/sign", "Mobile app icon variations" ] }, { "title": "Color Guide", "position": "Lower center left", "count": 5, "labels": [ "main color", "secondary color", "warm gray", "Light green", "accent color" ] }, { "title": "Font Standards", "position": "Shimochuchu", "count": 2, "labels": [ "Source Han Sans (CN)", "Source Han Sans Soft Black CN" ] }, { "title": "Minimum usable size", "position": "Lower Middle Right", "count": 2, "labels": [ "Minimum size for horizontal logo", "Minimum size for stacked logos" ] }, { "title": "Safe blank area", "position": "Bottom left", "count": 1, "labels": [ "Illustration of the safety margins around the logo" ] }, { "title": "Incorrect usage example", "position": "From bottom center to right", "count": 5, "labels": [ "No deformation allowed", "Color changes are prohibited", "No special effects allowed", "Vertical stretching is prohibited.", "Do not place on complex photo backgrounds" ] } ], "grid": "The layout features a rigorous multi-column format with fine dividing lines and consistent spacing." }, "logo_design": { "form": "GDX lettering", "integration": "The silhouette of a dog in negative space is embedded between G and X, forming both the shape of the letter D and the outline of a standing dog facing right.", "color": "Dark green", "subtitle": "The Chinese brand name is centered, with thin horizontal lines on both sides." }, "color_swatches": [ { "name": "Dark green", "hex": "#1E3D34" }, { "name": "off white", "hex": "#F5F3EF" }, { "name": "warm gray", "hex": "#E5E2DB" }, { "name": "Light green", "hex": "#A8C5B1" }, { "name": "warm orange", "hex": "#E0A86E" } ], "applications": { "mockups": 7, "details": "Business card sets, letter paper sets, smartphone app icon interfaces, website homepages with pet photos, dark green and kraft paper shopping bags, store signs, and combination square app icon variations." }, "typography": { "families": [ "Source Han Sans (CN)", "Source Han Sans Soft Black CN" ], "weights": [ "Bold", "Medium", "Regular" ] }, "rendering": "High-resolution 2D design showcases the project, including a front view, a clear vector logo, and realistic mockup embedding. Only subtle shadows are present within the product mockup itself, with no perspective distortion." }

{"type":"brand identity system board","brand":{"name":"{argument name=\"brand name\" default=\"狗东西\"}","english_mark":"{argument name=\"logo letters\" default=\"GDX\"}","industry":"{argument name=\"industry\" default=\"宠物行业\"}","date":"{argument name=\"date\" default=\"2024.05\"}","tagline":"爱它・懂它・陪伴它"},"style":{"overall":"minimalist premium corporate presentation, clean Swiss-inspired layout, large white margins, light gray divider lines, muted neutral background","palette":"deep forest green and warm off-white with small accent swatches","mood":"modern, trustworthy, pet-friendly, refined"},"canvas":{"aspect_ratio":"3:4 vertical","background":"soft warm white"},"header":{"left_title_cn":"品牌视觉识别系统","left_title_en":"BRAND IDENTITY SYSTEM","right_tagline":"爱它・懂它・陪伴它","center_logo":"large custom GDX wordmark with a white dog silhouette integrated into the middle of the logo, bold geometric curves and diagonal strokes, dark green","center_subtitle":"狗东西"},"layout":{"sections":[{"title":"基础信息","position":"upper left","count":3,"labels":["品牌名称","行业属性","设计时间"]},{"title":"设计网格","position":"mid upper left","count":1,"labels":["logo construction grid with ratio 1.618:1"]},{"title":"概念草图","position":"mid upper center","count":4,"labels":["dog head and letters sketch","outline refinement","intermediate combined mark","final simplified mark"]},{"title":"灵感来源","position":"mid upper right","count":6,"labels":["minimal white architectural arch","golden retriever profile photo","curved arch block composition","green leaf close-up","dark green material square","light wood texture square"]},{"title":"创意理念","position":"middle left","count":4,"labels":["设计哲学与符号意义","品牌定位服务","色彩心理学与策略","可扩展性与适应性"]},{"title":"品牌应用","position":"middle center to right","count":7,"labels":["名片 正反面","信纸信封","APP图标","网站首页 / 网站图标","产品包装 / 购物袋","店面门头 / 标识牌","mobile app icon variants"]},{"title":"色彩规范","position":"lower middle left","count":5,"labels":["主色","辅助色","暖灰色","浅绿色","强调色"]},{"title":"字体规范","position":"lower middle center","count":2,"labels":["思源黑体 CN","思源柔黑体 CN"]},{"title":"最小使用尺寸","position":"lower middle right","count":2,"labels":["horizontal logo minimum size","stacked logo minimum size"]},{"title":"安全留白区域","position":"bottom left","count":1,"labels":["clear space diagram around logo"]},{"title":"错误使用示例","position":"bottom center to right","count":5,"labels":["do not distort","do not change colors","do not add effects","do not stretch vertically","do not place on busy photo background"]}],"grid":"strict multi-column editorial board with thin dividing rules and consistent spacing"},"logo_design":{"form":"GDX lettermark","integration":"negative-space dog silhouette embedded between G and X, reading as the letter D while showing a standing dog profile facing right","color":"dark green","subtitle":"Chinese brand name centered below with thin horizontal lines on both sides"},"color_swatches":[{"name":"墨绿色","hex":"#1E3D34"},{"name":"米白色","hex":"#F5F3EF"},{"name":"暖灰色","hex":"#E5E2DB"},{"name":"浅绿色","hex":"#A8C5B1"},{"name":"暖橙色","hex":"#E0A86E"}],"applications":{"mockups":7,"details":"business card set, stationery set, smartphone app icon screen, website hero with dog photo, dark green and kraft shopping bags, storefront signage, grouped square app icon variations"},"typography":{"families":["思源黑体 CN","思源柔黑体 CN"],"weights":["Bold","Medium","Regular"]},"rendering":"high-resolution graphic design presentation board, flat front-facing view, crisp vector logo, realistic mockup inserts, subtle shadows only inside product mockups, no perspective distortion"}

Can you put{argument name="subject" default="My female dog"} Put the movie in{argument name="movie" default="Akira Kurosawa's film "Ran""} Is it correct?

A stunning and incredibly complex conceptual worldview infographic masterpiece, themed "{argument name="theme" default="The fundamental differences between Confucianism, Taoism and Buddhism"} The overall design is inspired by ancient Eastern mythological manuscripts. Background: Pure white vintage textured canvas with a light beige, aged parchment base, featuring slightly worn edges and water stain effects. Core Layout: The central visual employs a grand "vertical egg-shaped layered structure," with Buddhist, Taoist, and Confucian layers from top to bottom. Edge Details: The four corners are decorated with exquisite miniature illustrations themed around ancient observation notes, ritual objects, and runes. Color Scheme: Primarily low-saturation sage green, light gold, and off-white tones, creating a bright and soft overall effect without glaringly high-saturation colors. Details: Architectural lines, landscape brushstrokes, lotus patterns, and cloud layers are clearly visible, rendered with exceptional delicacy. Seamless Integration: The natural transition between the three layers is achieved through clouds and flowing water, seamlessly connecting the Buddhist light, Taoist Tai Chi cloud patterns, and Confucian scholarly atmosphere. Style:{argument name="art style" default="Classical ink painting line art + low-saturation watercolor by number"} It possesses a light and airy feel reminiscent of ancient Chinese manuscripts. Text annotations: Pure traditional Chinese characters, using a weathered ancient Song typeface. Each annotation includes a short title and a line of poetic description, connected to details by a thin, deep gold line, with no repetition. Aspect ratio: 3:4. Vertically independent and complete units. Title area (top): `Confucianism, Buddhism, Taoism: Fundamental Differences`. Central layered annotations: Top layer "Buddhism": `Buddhism`, `The Relationship Between Self and Self`, `Selflessness, Healing the Mind, Letting Go`. Middle layer "Taoism": `Taoism`, `The Relationship Between Self and All Things`, `Non-Action, Healing the Body, Humility`. Bottom layer "Confucianism": `Confucianism`, `The Relationship Between People`, `Selflessness, Governing the World, Taking Responsibility`. Side auxiliary annotations: Left side: `Purity: Clearing the Mind and Brightening the Eyes, Eliminating Troubles`; `Tranquility: Following Nature, Returning to One's True Self`; `Reverence: Reverence for Responsibility, Actively Engaging in the World`. Right side: "60+ years old: Cultivating the mind: Taking gains and losses lightly, returning to one's inner self"; "35-55 years old: Being a person: Balancing work and rest, following the rules"; "7-35 years old: Striving forward, achieving success". Bottom summary: "The balance between detachment and engagement with the world is the highest wisdom of life".

Use the user uploaded image as the only subject reference. Transform the person in the uploaded image into a realistic LEGO style minifigure while preserving their recognizable identity and outfit. Preserve the subject's facial likeness as much as possible within LEGO minifigure design limits, maintaining the same hairstyle and hair color adapted into a LEGO hairpiece, facial expression, clothing colors, clothing design and patterns, and any visible accessories. The result should clearly represent the same person translated into LEGO form. The character must follow authentic LEGO minifigure proportions, including a cylindrical yellow or skin tone LEGO style head, simple printed facial features adapted from the subject, a block shaped torso with printed clothing details, standard LEGO minifigure arms with curved hands, short LEGO minifigure legs, and a distinct molded plastic hairpiece matching the subject's hairstyle. Render the figure using realistic LEGO plastic material with a slight glossy sheen, subtle molded seams, and natural toy surface reflections. The face and torso should include printed details typical of LEGO minifigures, and the hairpiece should appear as molded plastic. The final image should be a high quality realistic 3D toy render with soft diffused studio lighting, subtle shadows that emphasize shape and plastic texture, and clean sharp focus with crisp detail. The framing must show the full body minifigure from head to feet in a centered composition with the entire figure fully visible. Use a plain white background with no scenery, no props, and no additional elements. The final style should feel like a photorealistic LEGO minifigure toy render with highly detailed plastic texture and studio lighting.

A top-down, realistic photograph of a white printed sheet of paper placed on a light-colored wooden desk showcases a computer-generated musical score. The page is horizontally centered with a slight sense of perspective, illuminated by soft natural indoor light with minimal shadows, resulting in a clean, printed document. At the top center of the page, the title is printed in a simple sans-serif font.{argument name="file name" default="kura.mid"} -{argument name="track name" default="CH05 Piano"} Below is a standard black staff system with a treble clef and 4/4 time signature, designed to resemble an automatically converted MIDI score rather than a finely typed score. On the far left of the first line of staff is the instrument name "CH05 Piano". The page should display five lines of horizontal staff, with the starting measure numbers on the left of each line being 18, 21, 25, 29, and 33 respectively. Near the first line of staff is a tempo marker indicating a quarter note equal to 120. The staff is filled with dense, somewhat stiff but logical black notes, accidentals, stems, slurs, and rests. The treble register contains numerous sharps and repeated short note combinations, visually resembling a MIDI-converted piano score. The first three lines of staff should be relatively long and almost fill the entire width; the fourth line should also be relatively long; the fifth... The lines should be short, occupying only the left side of the page. The sheet music must remain legible, but the spacing should have a mechanically generated appearance. Use clean white paper with wide margins, no handwriting, no colored elements, no extra graphics, and simply present a printed sheet music photographed on a desk.

A cinematic full-body portrait depicts a deeply unsettling humanoid alien standing alone in a dark, rocky canyon. Centered, facing the camera directly, its posture is neutral and upright, arms hanging naturally. The creature is extremely tall and slender, with long limbs, narrow shoulders, a sunken torso, clearly visible ribcage, taut muscles, long forearms, enormous hands with slender fingers, and long, digitigrade feet firmly planted on the damp, black ground. Its hairless, smooth, greyish-beige skin with a brownish, fleshy undertone possesses a semi-organic and biomechanical texture, with a moist sheen, leathery grain, stretched membranes, tendons, grooves, and subtle vascular stripes. The neck and upper chest merge into a disturbing organic structure with tentacle-like folds and open passages extending from the collarbone area upwards to the head. The head is broad, mushroom-shaped or hammer-shaped, with a ridged crown and outward-spreading lateral lobes, presenting a non-human silhouette; the face should appear elusive and exotic, rather than cute or familiar. Place the subject in a narrow canyon formed by dark, eroded rock, framed by two massive rocks on either side of the foreground, with a hazy, misty cliff in the background. Use cool, desaturated lighting, soft mist, damp surfaces, shallow depth of field atmospheric haze, and a somber monochromatic palette composed of charcoal, slate, ash, and soft flesh tones. The overall atmosphere should create a grotesque, realistic, oppressive, and biologically logical feel, like a still from a high-end science fiction horror film. Ultra-realistic style, clear creature textures, subtle depth of field effects, dramatic natural backlighting from above, no clothing, no tools, no background creatures, no blood, no text.

Using REFERENCE_0 and REFERENCE_1 as the FACS action-unit guide, create a photorealistic expression-control test sheet instead of an illustrated chart. Generate {argument name="grid layout" default="15 panels arranged in 3 rows and 5 columns"} showing the same {argument name="subject" default="realistic young female student"} repeated in every panel with identical framing, lighting, hairstyle, and {argument name="outfit" default="white school shirt with a gray ribbon tie"}. Apply a different subtle facial-expression change to each panel based on these {argument name="FACS action units" default="AU1, AU2, AU4, AU5, AU6, AU7, AU9, AU10, AU12, AU15, AU17, AU20, AU23, AU24, AU25"}, while preserving identity consistency and keeping the changes limited to the face. Use a close-up front-facing portrait crop from upper chest to top of head, natural outdoor backlight, shallow depth of field, and thin white dividers between panels. Do not include any chart titles, labels, AU text, explanations, or Japanese text.

A realistic, vertical promotional portrait is set against a warm, modern library or bookstore interior. A young East Asian woman stands behind a wooden counter, displaying a large book to the camera. She has long, straight, dark brown hair parted in the middle, wears delicate earrings, and appears neat, friendly, and professional. She wears a white shirt, a soft green striped tie, and a beige apron with a name tag reading "Librarian Ai-chan" and a small open book icon sewn on it. Her left hand holds the book upright against her chest, slightly tilted towards the viewer, while her right hand is open beneath the book, a classic product presentation pose. The book is the visual focus: its dark, glossy cover features a striking orange title, "Pollo AI" (a large line of text at the top, with a glowing line in the center), followed by a smaller Japanese subtitle and functional description, and a circular "P" logo in the lower left corner. To the left of the counter is a transparent acrylic nameplate with Japanese text above and the prominent "Ai-chan" logo below. To the right of the counter, three hardcover books are stacked horizontally, each with a numbered label and a Japanese title on its spine. Against a softly blurred background, a tall wooden bookshelf stands on the left, displaying books, small potted plants, and a warm-toned lamp; a decorative painting with Japanese text hangs on the upper right wall; and a potted green plant sits on the far right. A small white "Pollo.ai" watermark is added to the upper right corner. The image uses warm ambient lighting, a shallow depth of field, and primarily beige and wood tones to depict realistic skin and fabric textures. The composition is simple, giving it the feel of a professional lifestyle advertisement. The composition is a portrait from mid-thigh to head, in a 3:4 ratio.

A hyper-realistic screenshot of a macOS desktop showcases the workspace of a machine learning engineer at night. The image is taken from a frontal view, with a dark blue macOS menu bar at the top and the Dock visible at the bottom. Two main application windows are displayed side-by-side on the desktop. On the left is a dark-themed Visual Studio Code window occupying about two-thirds of the screen. The VS Code project, named "VISIONCLASSIFIER" in the file explorer sidebar, contains a realistic Python ML folder tree with 11 visible top-level or expanded items: .venv, data, raw, processed, images, notebooks, src, utils, config.yaml, requirements.txt, and README.md. Within the notebooks folder, two visible files are displayed: 01_data_exploration.ipynb and 02_model_training.ipynb. The src folder displays the actual ML code structure, including dataset.py, transforms.py, models, resnet.py, train, engine.py, trainer.py, and utils.py. Four tabs are open in the editor area: trainer.py, engine.py, resnet.py, and config.yaml, with trainer.py currently active. Clear and reliable Python training code for the ResNet image classification pipeline is displayed, including the Trainer class, train(self) and train_epoch(self, epoch: int) -> Dict[str, float] methods, referencing self.cfg.training.epochs, train_metrics, val_metrics, scheduler.step, save_checkpoint, self.model.train(), batch["image"], batch["label"], optimizer.zero_grad, criterion, loss.backward, optimizer.step, and accuracy(outputs, targets, topk=(1,))[0]. The code should be clear and have a natural screen feel, with line numbers displayed between lines 24 and 52. The VS Code window opens the integrated terminal's TERMINAL tab at the bottom, displaying the actual training logs for four epochs: Epoch 12/50, Epoch 13/50, Epoch 14/50, and Epoch 15/50. Each line contains training and validation data for Loss, Acc@1, and Acc@5, with the last line indicating that a new best checkpoint has been saved. The values ​​should reflect a successful training process, with Top-1 accuracy between 0.88 and 0.91, and Top-5 accuracy between 0.97 and 0.98. The bottom includes the standard VS Code status bar, displaying Python environment details. On the right is a dark-themed web browser window displaying a local dashboard on localhost:8000, titled "VisionClassifier | Dashboard," with the application title "VisionClassifier" and the subtitle "Image Classification Model." The dashboard comprises three stacked sections. The first section, "Model Overview," includes four metric cards: Top-1 Accuracy 91.23%, Top-5 Accuracy 98.30%, Total Parameters 23.51M, and Model ResNet-50. The second section, "Recent Training," displays a dark line graph of accuracy over 50 epochs, featuring two colored curves labeled Train (Top-1) and Val (Top-1), which rise rapidly and stabilize around 90%. The third section, "Confusion Matrix," displays a 10x10 heatmap with bright diagonal lines and axes labeled True and Predicted. Utilizing subtle reflections, clear typography, realistic UI spacing, and lifelike screen halo, the macOS top menu bar displays commonly used menus such as Code, File, Edit, Selection, View, Go, Run, Terminal, Window, and Help on the left, and system icons on the right, with the time displayed as Tue May 13 9:41 AM. The Dock should contain multiple recognizable application icons, giving an overall realistic and uncluttered feel. Overall style: hyper-realistic screenshot, professional developer workstation, refined dark mode interface, unstylized, without illustration-like elements, indistinguishable from a real screen screenshot.

A realistic half-length portrait, the subject being a{argument name="subject" default="young women"} Standing in bright natural sunlight, the photo is taken in a vertical format. She has long, straight hair.{argument name="hair color" default="Dark brown"} Her hair, glossy and lustrous, was parted in the middle and cascaded naturally over her shoulders. She wore a crisp white short-sleeved school uniform shirt with a navy blue striped tie, and a delicate embroidered badge on her left breast pocket featuring a crown and coat of arms. She stood at a slightly angled, relaxed posture, her head turned slightly to one side, creating a sophisticated, candid fashion sense. The background was set to a vibrant {argument name="background setting" default="Red Shrine Building"} Against a blurred background, rich red pillars and details of an arched blue roof are visible. Shallow depth of field and soft bokeh effects keep the subject sharp and in focus. The lighting should be clean, high-contrast, and embellished, with bright highlights on the hair and shirt to showcase realistic clothing folds and natural skin tones, presenting a refined, high-end portrait aesthetic similar to Japanese commercial photography.

A highly detailed, realistic anime-style portrait depicts a young woman crouching low and looking slightly downwards at the camera. She has long, flowing hair.{argument name="hair color" default="Gray Gold"} Her hair swayed gently in the breeze; her skin was fair; and her eyes were large and expressive. She wore {argument name="outfit" default="Japanese school uniform, paired with a light gray cardigan, white shirt, dark plaid bow tie, dark plaid pleated skirt, dark knee-high socks, and black leather loafers."} Her arms were casually draped over her knees. The background was bright.{argument name="sky condition" default="A clear blue sky, dotted with a few white clouds."} The bottom is faintly visible and blurry.{argument name="background setting" default="Barbed wire fence and green trees"} The lighting creates a campus atmosphere. The light is bright natural sunlight with soft, cinematic shadows, highlighting the realistic texture of clothing and skin.

One{argument name="photography style" default="Realistic portraits with shallow depth of field and soft bokeh effect"} The protagonist is{argument name="subject" default="Young Japanese women"} He was turning back to look at the camera, his face showing {argument name="expression" default="A gentle smile"} She was wearing{argument name="attire" default="Light beige kimono with orange maple leaf pattern"} A gold belt cinched her waist. Her long, dark hair was styled in an elegant updo, with a few strands falling beside her cheeks, and she wore delicate pearl earrings. (Background setting: [omitted]){argument name="setting" default="A garden ablaze with red leaves in autumn"} The upper left corner is adorned with vibrant red leaves, and the background is deeply blurred, creating a tranquil and cinematic atmosphere.

A photorealistic selfie showcases a{argument name="subject description" default="A young woman with a messy bun, dark skin, and brown hair."} Lying{argument name="location" default="The background features a beach, waves, and rocky cliffs."} On a striped beach towel. She is facing the camera.{argument name="facial expression" default="Winking and playfully sticking out her tongue."} She was wearing a white baseball cap and a white crew-neck swimsuit, both printed with bold black lettering.{argument name="brand text" default="ANTHROPIC"} She wore small gold hoop earrings and a delicate gold necklace. The bright, sunny light captured the relaxed, carefree atmosphere of summer, with blue sky and scattered clouds overhead.

A highly realistic Blu-ray disc cover features a portrait of a young Japanese woman. She has a messy black bob and wears large, round, thin-rimmed metal glasses. She wears a thick, textured, deep red knit sweater. Her expression is soft and captivating, looking directly at the camera. The background is a softly blurred, warm-toned interior environment, seemingly in a café, with a white coffee cup and tray faintly visible on the table to the right. The warm, cinematic lighting accentuates her facial features and the texture of her sweater. The image is framed within the standard blue border of a Blu-ray disc case, with the Blu-ray Disc logo prominently displayed at the top center, upper right, and lower left corners. On the left side are striking and elegant vertical Japanese characters…{argument name="main title" default="Witch's House Ringo"} Next to it is the phonetic alphabet "majo no ieringo", and it is decorated with delicate glowing pink lines and star/pentagram patterns. The lower right corner contains smaller Japanese characters "{argument name="subtitle" default="Like a witch, she will enchant you."} The letter “RINGO” and “MAJYONOIE” are printed below it.

Hyperrealistic interior photography works showcasing{argument name="aircraft model" default="Boeing 747"} The cockpit. The view is centered and perfectly symmetrical, positioned between the two pilot seats, showcasing the main instrument panel, central console, dual control sticks, and overhead switch panel at the top edge. The cockpit is empty; the two pilot seats are visible in the foreground, fitted with light gray or off-white sheepskin covers. Each control stick has a dark rectangular checklist or electronic note item clipped to it. The main panel is densely packed with realistic-looking avionics, analog dials, autopilot controllers, and radio panels, totaling six illuminated displays: two large central navigation displays, one smaller blue-toned flight display on each side, and two green text/data screens below the central console. The center contains four throttle levers, as well as additional flaps and system control sticks, rotary knobs, shielded switches, and the weathered brown-gray metallic surface typical of wide-body aircraft cockpits. The overhead panel should be covered with rows of switches, toggles, and indicator lights, some in shadow but clearly detailed. Through the windshield, a blurred airport gate scene is displayed, including the light-colored concrete tarmac, ground markings, and passenger boarding bridges connecting to or near the aircraft. The lighting is natural sunlight from outside, with soft lighting inside the cockpit, realistic shadows, subtle reflections on the glass screens, and documentary-style color gradation. It emphasizes extreme detail, realism, the true complexities of aerospace engineering, and image quality indistinguishable from real photographs. Shot with a full-frame camera, eye-level perspective, 24–28mm wide-angle lens, full depth of field on the instrument panel, high dynamic range, no people, no exaggerated stylization, and no illustration style.

A realistic indoor portrait, in which {argument name="subject" default="Elon Musk"} He stands before a traditional Chinese desk, practicing calligraphy on a large sheet of Xuan paper. He wears a black Tang suit with Chinese knot buttons and delicate embroidery, leaning slightly forward, one hand resting on the table, the other holding a brush suspended vertically above the paper. The scene is a warm and elegant scholar's study, furnished with dark wood furniture; to the left is a carved brush holder with four brushes hanging from it; to the left front is an inkstone and a small inkwell; a small porcelain vase sits on the side table behind him. Behind him hang two vertical scrolls: the left scroll prominently displays large characters.{argument name="left scroll text" default="Man Jiang Hong"} On the right, a scroll displays a long line of Chinese calligraphy. To the right is a wooden lattice window, letting in soft ambient light. The paper on the table is covered in flowing, semi-cursive black calligraphy, suggesting he is writing.{argument name="written poem" default="Man Jiang Hong"} The film employs a documentary-style aesthetic with photorealistic detail, combined with cinematic warm lighting, rich wood tones, authentic calligraphy tools, shallow depth of field, and half-body compositions to create a serious and focused atmosphere.

{ "type": "Photorealistic printed exam papers", "style": "Black text on slightly wrinkled off-white paper, realistic document photography effect.", "header": { "main_title": "{argument name=\"exam title\" default=\"Yulin City 2024 Spring Semester Grade 11 Final Exam Teaching Quality Monitoring Test Paper\"}", "subject": "{argument name=\"subject name\" default=\"Math Exam\"}", "details": "{argument name=\"score and time\" default=\"This exam paper has a total score of 150 points and the exam time is 120 minutes.\"}" }, "layout": { "instructions": { "title": "Precautions:", "point_count": 3, "content": "Standard exam rules regarding answer sheets, 2B pencils, and submitting the exam." }, "main_section": { "heading": "{argument name=\"section title\" default=\"I. Multiple Choice Questions: This section contains 8 questions, each worth 5 points, for a total of 40 points. For each question, choose the one option that best answers the question.\"}", "question_count": 8, "format": "A list of numbers from 1 to 8, each question containing a mathematical formula and four options (A, B, C, D).", "topics": [ "Set intersection", "plural", "Monotonically increasing function", "Necessary and sufficient conditions", "Vectors and Trigonometric Functions", "Solid geometry", "Lateral area of ​​a cone", "Trigonometric function graphs" ], "diagrams": { "count": 2, "descriptions": [ "The 3D wireframe of the cube next to question 6, with internal dashed lines.", "The two-dimensional Cartesian coordinate system diagram next to question 8 shows a partial sine wave curve." ] } }, "footer": "{argument name=\"footer text\" default=\"Grade 11 Mathematics Exam, Page 1 of 4\"}" } }

{ "type": "photorealistic printed examination paper", "style": "black text on slightly crumpled off-white paper, authentic document photography", "header": { "main_title": "{argument name=\"exam title\" default=\"玉林市 2024 年春季期高一期末教学质量监测试卷\"}", "subject": "{argument name=\"subject name\" default=\"数 学 试 卷\"}", "details": "{argument name=\"score and time\" default=\"本试卷满分 150 分，考试时间 120 分钟。\"}" }, "layout": { "instructions": { "title": "注意事项:", "point_count": 3, "content": "Standard exam rules regarding answer sheets, 2B pencils, and submission." }, "main_section": { "heading": "{argument name=\"section title\" default=\"一、选择题: 本题共 8 小题，每小题 5 分，共 40 分。在每小题给出的四个选项中，只有一项是符合题目要求的。\"}", "question_count": 8, "format": "Numbered list from 1 to 8, each with mathematical formulas and 4 multiple-choice options (A, B, C, D).", "topics": [ "Sets intersection", "Complex numbers", "Monotonically increasing functions", "Necessary and sufficient conditions", "Vectors and trigonometry", "Solid geometry", "Cone lateral surface area", "Trigonometric function graph" ], "diagrams": { "count": 2, "descriptions": [ "A 3D line drawing of a cube with internal dashed lines located next to question 6.", "A 2D Cartesian coordinate graph showing a partial sine wave curve located next to question 8." ] } }, "footer": "{argument name=\"footer text\" default=\"高一数学试卷 第 1 页 (共 4 页)\"}" } }

A realistic, high-resolution automotive photograph showcasing a car{argument name="car color" default="Bright red"} of{argument name="car model" default="Ferrari F8 Tributo"} Stop at{argument name="setting" default="Coastal highway overlooking the sea"} Above. The sports car is parked at a slight side angle, showcasing its sleek aerodynamic curves, aggressive front fascia, distinctive LED headlights, and silver alloy wheels with yellow center caps. The iconic yellow shield emblem is visible on the front fender. The background features the deep blue sea, low stone railings, and a distant rocky coastline covered in lush vegetation and scattered buildings under a clear blue sky. The lighting is {argument name="lighting" default="Bright sunny day"} It casts clear, realistic shadows on the asphalt road and creates a dazzling reflection on the car's glossy paint.

A realistic car photograph showcasing a vehicle{argument name="car color" default="Bright red"}{argument name="car model" default="Ferrari sports car"} Stop at{argument name="setting" default="Coastal Highway"} Above. The vehicle is angled towards the camera, showcasing its sleek aerodynamic curves, silver five-spoke alloy wheels, and sharp front fascia with black air intakes. The background is {argument name="background" default="Magnificent seascape with rocky coastline"} The deep blue sea and clear sky. Bright sunlight cast sharp shadows on the asphalt road, highlighting the glossy paint of the cars.

A photorealistic fashion magazine cover features a portrait of a beautiful young Asian woman. She has {argument name="hair style" default="Light brown hair with gold highlights and a braided crown"} Her makeup was soft and elegant, and her expression gentle. She wore a {argument name="clothing" default="White ribbed off-shoulder camisole top"} She wore delicate gold dangling earrings. The background was a soft, neutral gray. The layout featured elegant typography: a large masthead at the top, with the words "..." in a classic serif font.{argument name="magazine title" default="LUMINA"} The text on the left reads "THE NEW ERA OF AI BEAUTY," and below it is a larger text.{argument name="main headline" default="STYLE EVOLUTION"} The text on the right reads "FUTURE FASHION," and below it is "WHAT'S NEXT?". The text at the bottom reads "".{argument name="bottom text" default="BEAUTY SECRETS of TOMORROW"} Below is "ELEGANT & CHIC". The overall lighting effect is soft, professional studio lighting.

A photographic realism of modern{argument name="restaurant brand" default="BURGER KING"} The exterior of the fast food restaurant on a street corner, photographed from a left-front view, with a blue sky and white clouds in the background. The building features a beige stucco facade, with a striking red decorative strip along the roof edge. A large circular brand logo is located on the upper left of the front wall, and a huge red three-dimensional illuminated lettering spans the top, spelling out [the logo's name].{argument name="main sign text" default="BURGER KING"} On the right side of the facade, a tall vertical menu billboard displays six clearly visible product panels and prices: Whopper $6.99, Plant-based Whopper $7.49, Chicken and Fries $3.49, Onion Rings $2.49, two $6 combo panels, and a $3 all-you-can-eat combo panel including fries and a drink. On the left is a black drive-thru canopy with the text "..."{argument name="drive thru sign" default="DRIVE THRU"} There are two cars in the driveway: a dark SUV close to the camera and a silver sedan ahead. The front corner should have floor-to-ceiling windows, revealing a bright interior environment, with several customers and menu items vaguely visible. A recruitment poster is placed on the front window, with the text {argument name="window poster text" default="Now Hiring"} The building's edges are meticulously landscaping, with five shrubs planted in dark-covered flowerbeds. The foreground features a concrete sidewalk, lawn, utility poles, traffic lights, and a suburban intersection extending to the left rear. Natural midday light, crisp shadows, a commercial real estate photography style, ultra-clean signage, highly legible text, realistic reflections on the glass, sharp architectural details, and a subtle white accent in the upper right corner.{argument name="watermark text" default="Pollo.ai"} Watermark.

A photorealistic portrait, the subject is{argument name="subject description" default="A young East Asian woman"} Sitting{argument name="setting" default="Dimly lit izakaya"} She stood in front of the wooden bar counter. She had long, black, curly hair and wore {argument name="top clothing" default="Dark red silk shirt"} She wore a black mini-skirt, black sheer pantyhose, and red high heels. She was holding {argument name="drink" default="A beer"} She pouted slightly and looked at the camera. Blurred customers and warm colors could be seen in the background.{argument name="lighting source" default="Paper lanterns"} Cinematic lighting with a shallow depth of field effect.

{ "type": "Photorealistic front pages of major Japanese newspapers", "style": "Hyperrealistic, shot from a slightly overhead angle and placed on a flat surface, with natural light and clear typography.", "header": { "date": "{argument name=\"publication date\" default=\"Thursday, April 23, 2026 (Reiwa 8)\"}", "title": "{argument name=\"newspaper name\" default=\"Asahi Shimbun\"}", "sponsor_ad": "~Living with Water~ SUNTORY" }, "layout": { "headlines": [ "{argument name=\"main headline\" default=\"OpenAI releases new model "Spud"\"}", "{argument name=\"secondary headline\" default=\"The "Codex SuperApp" will be released simultaneously.\"}" ], "sub_banner": "It comprehensively surpasses existing benchmarks in terms of network attack and coding performance.", "centerpiece_image": { "description": "Dark blue background with the text \"{argument name=\"embedded image subject\" default=\"Codex SuperApp\"} \"and three mobile UI screens displaying application interfaces.\"", "caption": "Codex SuperApp concept art (provided by OpenAI)" }, "vertical_side_text": [ "\"To become the cornerstone of the AGI era\"", "Development competition intensifies further" ], "article_sections": [ { "sub_headline": "Experts gave it high praise, calling it \"overwhelming progress\".", "bullet_points_count": 3, "bullet_points": [ "The ability to automatically detect and respond to cyberattacks has improved by 4.2 times compared to the past.", "Break the SWE-bench high score in coding-assisted benchmark tests.", "Natural language understanding and reasoning abilities have been greatly enhanced." ], "body_text": "The main body of the article is written in dense Japanese, discussing the release of AI models and quoting Sam Altman." } ], "bottom_teasers": { "count": 3, "items": [ "Tariff Negotiations: Japan and the US to Renegotiate (Economics Section, Page 3)", "Heavy rains hit Kyushu region, injuring two people (Social Affairs section, page 28)", "Opinion: How to Deal with AI" ] } } }

{ "type": "Photos of the front page of a printed newspaper", "style": "Photorealistic, natural light, slightly angled perspective, printed on textured newsprint.", "header": { "left_ad": "SHINSEGAE DUTY FREE logo", "title": "{argument name=\"newspaper name\" default=\"South Korean economy\"}", "right_ad": "Myungryon Jinshi Spare Ribs (명륜진사갈비) logo and small image of the meat", "date_and_info": "{argument name=\"publication date\" default=\"Thursday, April 23, 2026\"}" }, "headlines": { "main": "{argument name=\"main headline\" default=\"OpenAI releases next-generation model 'Spud', sweeping coding and cyber warfare benchmarks.\"}", "sub": "The simultaneous launch of the new super app 'Codex'... Could it become a game-changer in the \"AI war\" landscape?" }, "layout": { "centerpiece": { "type": "Promotional image", "title_text": "Introduced with great fanfare{argument name=\"app name\" default=\"Codex\"} A super app for everyone", "visuals": "3 smartphones displaying dark mode UI", "phone_1": "The left side of the phone displays the function menu list.", "phone_2": "The middle mobile phone displays an analysis line chart.", "phone_3": "The phone on the right displays a code editor with Python scripts." }, "sections": [ { "title": "Left column", "position": "center left", "count": 2, "labels": [ "Introduction Article Text", "First in all coding benchmark tests" ] }, { "title": "Right-side column", "position": "center right", "count": 1, "labels": [ "They also achieved a resounding victory in cyber warfare benchmark tests." ] }, { "title": "Article in the bottom left corner", "position": "Bottom left", "count": 1, "labels": [ "New York stocks rose across the board, driven by gains in AI-related stocks... The Nasdaq index rose 2.7% ↑" ] }, { "title": "bottom right corner ad", "position": "Bottom right", "count": 1, "labels": [ "{argument name=\"bottom ad text\" default=\"Samsung Electronics, the world's leading memory semiconductor company\"}" ] } ] } }

{ "type": "photograph of a printed newspaper front page", "style": "photorealistic, natural lighting, slightly angled perspective, printed on textured newsprint paper", "header": { "left_ad": "SHINSEGAE DUTY FREE logo", "title": "{argument name=\"newspaper name\" default=\"한국경제\"}", "right_ad": "명륜진사갈비 logo with a small image of meat", "date_and_info": "{argument name=\"publication date\" default=\"2026년 4월 23일 목요일\"}" }, "headlines": { "main": "{argument name=\"main headline\" default=\"오픈AI, 차세대 모델 'Spud' 공개 코딩·사이버전 벤치마크 석권\"}", "sub": "신규 슈퍼앱 'Codex'도 출시… “AI 전쟁” 판도 바꿀 게임 체인저 될까" }, "layout": { "centerpiece": { "type": "promotional graphic", "title_text": "Introducing {argument name=\"app name\" default=\"Codex\"} The Super App for Everyone", "visuals": "3 smartphones displaying dark mode UI", "phone_1": "left phone showing a feature menu list", "phone_2": "middle phone showing an analytics line chart", "phone_3": "right phone showing a code editor with a python script" }, "sections": [ { "title": "left column", "position": "mid-left", "count": 2, "labels": ["introductory article text", "코딩 벤치마크 모두 1위"] }, { "title": "right column", "position": "mid-right", "count": 1, "labels": ["사이버전 벤치마크도 압승"] }, { "title": "bottom left article", "position": "bottom-left", "count": 1, "labels": ["뉴욕증시, AI 랠리에 일제 상승…나스닥 2.7% ↑"] }, { "title": "bottom right ad", "position": "bottom-right", "count": 1, "labels": ["{argument name=\"bottom ad text\" default=\"세계 1위 메모리 반도체 기업 삼성전자\"}"] } ] } }

A photographic, realistic nighttime street scene, showcasing the bustling activity.{argument name="city location" default="Tokyo"} The street was illuminated by a dense array of vertical neon signs and billboards. On the left was a striking bright pink sign that read, "..."{argument name="pink sign text" default="カラオケ"} Next to it was a white sign that read "Gyutaku Yakiniku" (牛角焼肉). The prominent sign on the right included a yellow {argument name="yellow sign text" default="Manga Cafe 4F"} ", a white "Southeast Main Building 2F", and a red "{argument name="red sign text" default="Chung Hwa Restaurant Ichibankan B1"} In the distant center, a huge blue billboard displays {argument name="billboard text" default="open TOKYO"} Above it is a digital screen displaying images of people with the words "YUNIKA VISION" on it. The street is crowded with the silhouettes of pedestrians as they move through this vibrant, light-and-shadow-filled urban canyon. The contrast between light and shadow is striking, and the rich neon colors are reflected in the surrounding environment.

A photorealistic night street photography shot of a bustling {argument name="city location" default="Tokyo"} street, heavily illuminated by densely packed vertical neon signs and billboards. On the left, a bright pink sign prominently reads "{argument name="pink sign text" default="カラオケ"}" next to a white sign reading "牛角 焼肉". On the right, prominent signs include a yellow one reading "{argument name="yellow sign text" default="まんが喫茶 4F"}", a white one reading "東南 本館 2F", and a red one reading "{argument name="red sign text" default="中華食堂 一番館 B1"}". In the center distance, a large blue billboard displays "{argument name="billboard text" default="open TOKYO"}" above a digital screen showing a person with the text "YUNIKA VISION". The street level is crowded with the backs of pedestrians' heads walking through the vibrant, glowing urban canyon. The lighting is high contrast with rich neon colors reflecting off the surroundings.

{ "type": "Photorealistic physical newspapers on the counter", "environment": "A casual indoor kitchen setting, with visible stove rim, natural lighting, and slightly curved paper with realistic creases.", "newspaper_header": { "logo": "A large blue circle, next to which is{argument name=\"newspaper name\" default=\"USA TODAY\"}", "date": "{argument name=\"date\" default=\"04.23.26\"}", "subtext": "National News | $3.00 | Thursday", "top_right_teaser": { "image": "naval vessels", "headline": "The United States and the Philippines held joint exercises in the South China Sea." } }, "layout": { "total_articles": 4, "total_embedded_images": 4, "sections": [ { "position": "Centered at the top", "headline": "{argument name=\"main headline\" default=\"OpenAI's new Spud model dominates in cyber warfare and coding benchmarks.\"}", "author": "Author: Jefferson Graham", "image": "{argument name=\"main image subject\" default=\"A 3D rendered blue whale wearing a crown, bearing the inscription 'spud by OpenAI'.\"}", "columns": 2 }, { "position": "center right", "headline": "Stocks rose, boosted by earnings reports from tech companies.", "author": "Author: Nathan Bomey", "columns": 1 }, { "position": "Bottom left", "headline": "{argument name=\"bottom headline\" default=\"The Codex app was released as a 'super app for all'.\"}", "author": "Author: Jessica Guynn", "image": "A smartphone screen displaying an application interface with the text 'codex'.", "columns": 2 }, { "position": "Bottom right", "headline": "New bird species discovered in the Amazon", "author": "Author: Doyle Rice", "columns": 1 }, { "position": "footer", "image": "rugby players in the game", "text": "NFL Draft Shocks: Quarterback Falls to Second Round" } ] } }

{ "type": "Photorealistic images of the front page of a printed newspaper", "setting": "Placed on the kitchen countertop, with blurred backgrounds of appliances.", "newspaper_header": { "title": "{argument name=\"newspaper name\" default=\"USA TODAY\"}", "date": "{argument name=\"date\" default=\"04.23.26\"}", "details": "National News | $3.00 | Thursday", "top_right_teaser": { "headline": "The United States and the Philippines held joint exercises in the South China Sea.", "image": "Naval vessels in the ocean" } }, "layout": { "element_counts": { "articles_and_teasers": 5, "printed_images": 4 }, "sections": [ { "position": "Centered at the top", "type": "Main article", "headline": "{argument name=\"main headline\" default=\"OpenAI's new Spud model dominates in cyber warfare and coding benchmarks.\"}", "image": "{argument name=\"main image subject\" default=\"A 3D blue whale wearing a crown, bearing the words 'spud by OpenAI'.\"}", "columns": 2 }, { "position": "Top of right sidebar", "type": "Side article", "headline": "Positive earnings reports from tech stocks boosted the stock market.", "columns": 1 }, { "position": "Centered at the bottom", "type": "Secondary Articles", "headline": "{argument name=\"secondary headline\" default=\"The Codex app was released as a 'super app for all'.\"}", "image": "Smartphones displaying an app interface with the word 'codex' on it", "columns": 3 }, { "position": "bottom of right sidebar", "type": "Side article", "headline": "New bird species discovered in the Amazon", "columns": 1 }, { "position": "bottom edge", "type": "Footer preview", "headline": "NFL Draft Shocks: Quarterback Falls to Second Round", "image": "American football player" } ] }, "photography_style": "Viewed from above in natural light, the paper texture and slight creases are visible." }

A realistic portrait of an elderly couple sitting opposite each other, both smiling.{argument name="setting" default="Cozy traditional tavern"} Beside a simple wooden table inside. The man on the left has white hair and a thick white beard, and is wearing a {argument name="man's clothing" default="Olive green fleece jacket"} The woman on the right has short gray hair, wears glasses, and is dressed in {argument name="woman's clothing" default="Blue denim jacket and white scarf"} They all looked directly at the camera, each holding a glass.{argument name="beverage" default="Pint beer"} There was a {argument name="table snack" default="Potato Chips"} A small, vintage lantern is lit. The background features warm, inviting lighting and a shallow depth of field effect, blurring the wood-paneled wall, framed photos, and several other customers.

A highly detailed and realistic portrait of a young person{argument name="ethnicity" default="Japan"} A portrait of a woman. She kept it.{argument name="hair style" default="Long black hair with bangs"} She had fair, flawless skin and large, deep brown eyes. She was wearing a {argument name="clothing" default="Simple white crew neck shirt"} noodles{argument name="expression" default="A gentle smile"} The camera is directly facing the lens. The light is soft and textured, creating a cinematic atmosphere with a shallow depth of field. The background is {argument name="background" default="Warm, blurry light spots"} .

A photorealistic {argument name="animal" default="raccoon"} dressed as a French painter, wearing a black beret and paint-stained smock, painting a portrait of a {argument name="portrait subject" default="seal"} on a canvas easel. Set on the {argument name="location" default="Champs-Élysées in Paris"} with the Eiffel Tower in the background, sunny day, trees lining the avenue.

A highly detailed, photorealistic portrait depicting a{argument name="subject description" default="Beautiful young Asian woman"} Leaning gracefully{argument name="furniture" default="White modern sofa"} She was wearing a {argument name="clothing" default="White silk spaghetti strap mini dress"} ,{argument name="hair style" default="Dark brown long curly hair"} Her arms draped softly over the white pillow. Her posture was relaxed and intimate, one arm gracefully raised above her head, the other gently resting on her abdomen, her gaze fixed on the camera, her expression gentle and captivating. The scene was created by {argument name="lighting style" default="Soft, natural sunlight streaming through the window"} The light casts soft, diffused shadows on flawless skin and pristine white cushions. The overall aesthetic is bright, minimalist, and ethereal, shot with an 85mm lens to achieve a cinematic shallow depth of field and soft, luminous highlights.

A realistic selfie, the main character is{argument name="subject description" default="A young Japanese woman"} Keep it{argument name="hair style" default="Long, wavy brown hair with airy bangs"} She was wearing a{argument name="clothing" default="White textured shirt"} She wore a V-neck dress and a delicate silver necklace, smiling gently at the camera. Her arm was outstretched, holding the camera in a typical selfie pose. The background was {argument name="location" default="Shibuya Streets"} A bright daytime scene. To her left, there is a realistic Japanese street address sign, clearly stating "..."{argument name="street sign text" default="1-8 Jinnan, Shibuya-ku"} The background shows a pedestrian crossing, pedestrians, city buildings, and a prominent curved building with the word "MODI" on top, next to a large digital billboard. The lighting is natural and soft, capturing a relaxed and everyday urban atmosphere.

{ "type": "Screenshot of a mobile short video app", "style": "Photorealistic quality, highly detailed", "main_image": { "subject": "{argument name=\"subject description\" default=\"A young Asian woman with long dark hair, wearing a white top, turned to look at the camera.\"}", "lighting": "Golden hour, sunset backlight, warm and soft light", "background": "{argument name=\"background location\" default=\"At sunset, the riverbank is framed by a skyline of towering, distorted skyscrapers.\"}" }, "ui_overlay": { "top_bar": { "left": "Time: 9:41, menu icon", "center_tabs": [ "experience", "Group buying", "focus on", "recommend" ], "right": "Signal strength, WiFi, 100mAh battery, search icon" }, "right_sidebar": { "count": 6, "elements": [ "Avatar with a red plus button", "Heart-shaped icon with the text '{argument name=\"like count\" default=\"236,000\"} '", "Comment bubble icon with text '16,000'", "The star icon displays the text '28,000'.", "Share arrow icon with text '37,000'", "Rotating record icon" ] }, "bottom_left_info": { "username": "{argument name=\"username\" default=\"@LittleDeer'sDailyLife\"}", "caption": "{argument name=\"caption\" default=\"Feeling the evening breeze and watching the sunset, this moment is so therapeutic~ 🌅\"}", "audio_track": "🎵 Original soundtrack from @XiaoLu's Daily Creations" }, "bottom_nav": { "count": 5, "tabs": [ "front page", "Friend (marked with a red badge #2)", "Central square plus button", "Message (with red badge #5)", "I" ], "footer": "White homepage indicator bar" } } }

A realistic snapshot of a restaurant shows a young Japanese woman sitting at a traditional Japanese restaurant's wooden sushi counter in warm lighting, positioned in the center of the frame and facing the camera. She has long, straight hair.{argument name="hair color" default="black"} Her hair was styled with bangs, and she wore a soft beige knitted off-the-shoulder sweater with a wide, sailor-style neckline trimmed in brown, over which was a white button-down collar. The long sleeves had flared cuffs, showcasing a feminine and elegant style. She held chopsticks in her right hand, hovering above a small blue dish, while her left hand rested on the table. In front of her was a rectangular ceramic plate neatly arranged with eight pieces of sushi: pink tuna, dark red tuna, orange salmon, two pieces of shrimp nigiri sushi, and pale white fish. To her right was another rectangular plate containing assorted tempura, including two large shrimp and mixed vegetable tempura. To the left of the table were a ceramic teacup with striking black Japanese calligraphy, a wooden chopstick holder filled with disposable chopsticks, and some small packages. Behind her was an open sushi bar, where two chefs in white uniforms and chef hats could be seen, along with a glass display case, a handwritten Japanese menu, a wooden bar counter, and several customers dining in the background. Shot from eye level, it has the quality of smartphone photography, presenting a natural indoor perspective, a documentary style, rich wood texture, warm amber lighting, delicate food details, and shallow to medium depth of field, creating a real and busy dinner atmosphere. It is unstylized, not an illustration, and has a photorealistic feel.

{ "type": "Photorealistic photography of printed newspapers", "style": "Realistic paper texture, slightly angled top view, natural lighting effects.", "header": { "left_logo": "SHINSEGAE DUTY FREE", "center_title": "{argument name=\"newspaper name\" default=\"South Korean economy\"}", "right_logo": "Myeongryun Jinsa Galbi", "date": "{argument name=\"date\" default=\"Thursday, April 23, 2026\"}" }, "headlines": { "main": "{argument name=\"main headline\" default=\"OpenAI releases next-generation model 'Spud', sweeping coding and cyber warfare benchmarks.\"}", "sub": "The brand-new super app 'Codex' has been launched... Will it become a game-changer in the AI ​​war?" }, "layout": { "sections": [ { "title": "Central graphic", "position": "middle", "text": "Introduced with great fanfare{argument name=\"app name\" default=\"Codex\"} A super app for everyone", "ui_screens_count": 3, "ui_screens_labels": [ "A menu list containing options such as writing code and data analysis.", "Analysis dashboard with purple line graph", "A code editor that displays Python Fibonacci functions." ] }, { "title": "Article column", "position": "Around the central graphic", "count": 3, "labels": [ "The left column containing the article text", "The middle column below the graphic", "Right column with key points" ] }, { "title": "bottom area", "position": "bottom", "headline": "{argument name=\"bottom headline\" default=\"Driven by a surge in AI-related stocks, New York stocks rose across the board... the Nasdaq index climbed 2.7%.\"}", "ad_box": "The text in the dark blue box reads: Samsung Electronics, the world's leading memory semiconductor company." } ] } }

A realistic studio portrait, the subject is{argument name="subject ethnicity and gender" default="A young Asian woman"} She kept it.{argument name="hair style" default="Long, dark, wavy hair"} Her face was adorned with faint freckles, and fine beads of sweat glistened on her skin, giving her a fresh and sporty appearance. She wore {argument name="outfit" default="An all-white tennis outfit, paired with a fitted short-sleeved polo shirt and a pleated mini skirt."} She casually slung a gun over her shoulder.{argument name="prop" default="White tennis racket"} He looked calmly directly into the camera. The background was {argument name="background" default="Seamless bright white studio background"} It is also equipped with soft, high-key lighting.

A lifelike close-up portrait of a beautiful young woman.{argument name="hair style" default="Golden long waves"} Her hairstyle, coupled with her captivating blue eyes, is a gentle and warm smile gazing at the camera. She is dressed in warm clothing and wearing {argument name="winter outfit" default="A white ribbed knit beanie with pom-poms, paired with a thick white turtleneck sweater in the same color family, and a beige plaid winter coat."} Soft snowflakes drifted gently around her, a few landing naturally on her hair, hat, and shoulders. The background was blurred.{argument name="background" default="Snow Forest"} In the top left corner, there is a dark, semi-transparent rounded rectangle containing the white text "{argument name="overlay text" default="GPT Image-2"} ".

Generate a physical map{argument name="subject" default="High school exams"} topic{argument name="type" default="Multiple choice questions"} The image is at 9:16. Generate 4 images, each corresponding to one answer to the given question.