GPT Image 2 prompts

Create an ultra-cinematic dark fantasy action illustration in a wide 16:9 frame, like a high-budget game key visual. Show exactly 4 humanoid figures in a chaotic aerial battle: 1 central silver-haired male warrior lunging toward the viewer with one arm outstretched, wearing ornate black-and-silver armor with gold details, intense focused expression, long hair blown upward by force; 1 pale red-haired wounded figure on the left reclining or falling backward in flowing white fabric, face tilted upward, scratched and bloodied; 1 enormous pale white-haired face in extreme close-up on the right edge, partially cropped, eyes shadowed, with blood marks and strands of hair filling the frame; and 1 small distant fiery red-haired fighter near the upper center, surrounded by branching golden lightning. The scene should feel like {argument name="fantasy battle theme" default="a clash of immortal warriors in a storm of fire, lightning, and wind"}. Use {argument name="main character hair color" default="silver-white"} for the central warrior, {argument name="armor style" default="ornate black-and-silver fantasy armor with gold accents"}, and {argument name="energy color" default="golden lightning and orange fire"}. Compose with extreme dynamic perspective, motion blur, swirling hair and cloth, embers, smoke, sparks, shattered debris, glowing highlights, and warm firelight against cool gray shadows. Make the image painterly yet hyper-detailed, dramatic, emotional, and epic, with no text, no logos, no UI, and no watermark.

Goal: Create a realistic dark-mode social media post screenshot about a leaked AI image model, styled like an X/Twitter post.

Canvas: Square 1024×1024 image, black background, with a single tweet occupying the full frame. Use authentic X/Twitter dark mode spacing, white text, blue hashtag/link color, and slightly rounded embedded media corners.

Tweet header: Top left circular profile photo of a light-skinned middle-aged man in a white shirt. Display name is {argument name="display name" default="White"} in bold white text with a blue verified checkmark, and handle {argument name="username" default="@White1637402"} in gray beneath or beside it. Top right has two small UI icons: a stylized X/Twitter action icon and a three-dot menu.

Tweet text: Use the exact visible post copy: “A new GPT image model just appeared on Arena called mona-lisa-1. It's extremely better than Image 2. 💪💪” followed by a blank line and the blue hashtag {argument name="hashtag" default="#ChatGPT"}. Keep line breaks and spacing like a real tweet.

Embedded media layout: Below the tweet text, show exactly 2 side-by-side image cards in one row, separated by a thin black gutter, both with rounded outer corners.

Left media card: A glossy 3D illustration of a cute white-and-black robot mascot in a classroom or chalkboard setting, waving with its right hand and holding a paper sign. The robot has a rounded black face screen with two glowing cyan smiling eyes, white helmet-like head shell, white body panels, black joints, and a friendly pose. Behind it is a chalkboard covered with handwritten educational text and doodles. Large chalk text reads “Hi there! I’m ChatGPT” on the left, and “That’s me!” with an arrow pointing toward an OpenAI swirl logo on the upper right. The robot’s paper sign reads “I was created by OpenAI” with smaller text underneath. Add realistic 3D lighting, slight depth of field, and toy-like materials.

Right media card: A dramatic cinematic poster-like fantasy/business scene showing exactly 3 human figures: two giant suited men looming in the stormy cloudy sky above, and one small caped figure seen from behind in the foreground. The two giant men resemble tech executives, wearing dark business suits and looking down sternly; one has glasses and gray hair, the other has curly hair. The foreground figure wears a bright royal-blue cape with a white logo and the word “claeposeok” or similarly misspelled text on the back, standing on rocky ground and holding a sword at their side. The sky is dark, smoky, and epic with strong rim lighting, blue-gray clouds, and heroic scale.

Visible text count: Include exactly 7 main visible text groups: 1 display name “White”, 2 handle “@White1637402”, 3 tweet body sentence, 4 hashtag “#ChatGPT”, 5 chalkboard phrase “Hi there! I’m ChatGPT”, 6 chalkboard phrase “That’s me!”, 7 paper sign “I was created by OpenAI”. Minor chalk scribbles may be present but should be small and not dominant.

Visual style: Photorealistic screenshot compositing with crisp UI text, realistic embedded image thumbnails, high contrast, and no extra watermark. Keep the overall look like a viral tech leak tweet screenshot discussing {argument name="model name" default="mona-lisa-1"}.

Create a high-fashion studio portrait of a {argument name="subject" default="young adult woman"} with delicate natural facial features, a cute subtle mixed-heritage look, soft face shape, dimensional features without an overly influencer-like appearance, and long voluminous layered {argument name="hair color" default="black hair"} with slightly messy lazy flyaway strands; several loose strands fall naturally across her cheeks and forehead with realistic hair texture. She sits on a seamless warm off-white studio floor and backdrop, posed low on the ground with one leg bent high toward the camera, the other leg folded and extended diagonally, one hand supporting her behind her, shoulders slightly hunched inward, gazing directly at the camera with a calm, slightly sultry expression. Outfit: fitted black long-sleeve top with a low square neckline, black sparkling sequined mini skirt, sheer black stockings, and black ankle-strap stiletto sandals. Use soft diffused editorial lighting, realistic skin texture, muted neutral tones, shallow but not extreme depth of field, full-body vertical composition with generous negative space, magazine lookbook styling, no harsh shadows, no watermark. Add exactly 3 small minimalist typography blocks in tiny uppercase serif text: top left reads “I’M NOT / A REBEL, / I’M JUST NOT YOU.”; lower left beside a short horizontal line reads “BEAUTY IS / POWER; / A SMILE IS / ITS SWORD.”; right middle reads “TOO SOFT / FOR ALL OF IT.” with a thin vertical bracket mark. Keep the typography subtle, black, and very small against the blank background.

Goal: Create a clean studio character consistency sheet for {argument name="character name" default="a young woman"}, designed as a model reference board for gpt-image-2.

Canvas: Wide horizontal 16:9 white background, minimalist editorial layout, soft even studio lighting, realistic high-fashion photography style, sharp but natural skin texture, no watermark.

Layout: Use exactly two rows. The top row contains 5 panels: 3 full-body standing turnarounds on the left labeled FRONT, SIDE, and BACK, plus 2 cropped detail callout panels on the right. The bottom row contains exactly 6 square head-and-shoulders expression portraits aligned evenly across the width.

Subject details: The same young East Asian woman appears consistently in every panel, with pale skin, slim build, soft oval face, large eyes with subtle eyeliner, straight brows, small nose, full natural lips, and {argument name="hair color" default="messy dusty pink hair"}. Her hair is long, slightly wavy, tousled, side-parted, and voluminous, falling past the shoulders. Outfit is consistent: {argument name="top" default="loose gray off-shoulder knit sweater"}, white straight-leg pants, and white sneakers. Keep the sweater slouchy with one shoulder exposed and long sleeves covering part of the hands.

Top row details: The 3 full-body views show the character standing neutrally against pure white: FRONT view facing camera, SIDE view in profile, BACK view turned away with hair visible down the back. Add small uppercase black labels above each full-body figure: FRONT, SIDE, BACK. To the right, include a close-up face portrait with a thin black horizontal callout line pointing to the hair and the label MESSY PINK HAIR. Beneath it, include a cropped shoulder-and-knit texture detail with a thin black callout line and the label OFF-SHOULDER KNIT.

Bottom row expression set: Include exactly 6 expression portraits of the same character, each cropped from chest or shoulders up and labeled in small uppercase black text underneath. From left to right: 1) NEUTRAL, calm mouth closed; 2) SMILE, bright friendly smile showing teeth; 3) SERIOUS, focused slightly pouty expression; 4) SURPRISED, eyes wide and mouth open; 5) SAD, subdued downturned mouth and tired eyes; 6) ANGRY, furrowed brows and tight pout.

Visual style: High-key white studio photography, clean fashion look, crisp product-reference composition, consistent face identity across all panels, realistic proportions, soft shadows, muted pastel pink and gray color palette.

Constraints: Use exactly 3 full-body views, exactly 2 detail callouts, and exactly 6 expression portraits. Keep all labels in English uppercase, small black sans-serif type. Do not add extra props, extra characters, logos, borders, decorative graphics, or additional expressions.

Create an epic vertical fantasy xianxia scene showing {argument name="character name" default="a lone immortal swordsman"} standing on the jagged peak of a dark mountain, viewed from behind in a heroic low-angle composition. The warrior wears flowing {argument name="robe color" default="white and silver"} ancient Chinese robes with long wind-whipped sleeves, layered fabric, ornate metallic trim, a tied waist sash, and long black hair streaming in the storm. He raises one straight sword vertically toward the heavens, and the blade becomes the centerline of a blinding golden-white beam of energy that splits open a colossal heavenly gate or stone door in the sky. The gate is ancient, monumental, and symmetrical, made of two enormous dark stone slabs covered in intricate circular carvings, runes, relief patterns, and gold filigree, cracking apart down the middle under the sword strike. Surround the split with intense lightning, arcing plasma, sparks, dust, glowing fissures, and levitating chunks of broken rock; count the visible main elements as 1 swordsman, 1 raised sword, 1 central light beam, 2 separated gate halves, and many scattered debris fragments. The setting is above storm clouds at dusk, with towering dark clouds, golden backlight, dramatic volumetric rays, thunderbolts, mist, and a sense of divine scale. Include distant silhouettes of traditional Chinese mountain temples and pagoda rooftops on the lower left and lower right edges, partly swallowed by clouds. Use cinematic fantasy concept art style, ultra-detailed, high contrast, vertical 3:4 composition, sharp focus on the sword and radiant fissure, atmospheric perspective, dark teal-gray storm tones contrasted with molten gold light, awe-inspiring scale, no modern objects, no text, no watermark.

Create a bright, high-detail anime portrait of a cheerful young woman in a summery street-fashion outfit. She has very long flowing {argument name="hair color" default="platinum white"} hair with soft windblown strands, large sparkling blue eyes, fair skin, a blush, and an open smiling expression. Add an elegant floral hair arrangement on the right side of her head made of cream-white blossoms and small baby’s-breath sprigs, plus delicate dangling flower earrings. She is posed in a lively three-quarter view, upper body to mid-thigh, one hand lifted near her cheek as if waving or playfully posing, looking directly at the viewer. Dress her in a {argument name="top color and pattern" default="dark charcoal off-shoulder blouse with large white polka dots"}, with ruffled elastic trim around the neckline and sleeve cuffs, slightly loose fabric, and a small coral-red ribbon accent near the chest. Pair it with a {argument name="skirt style" default="white cargo mini skirt"} featuring visible patch pockets, seams, belt loops, and a clean fitted waist. The background is an outdoor sunny city street rendered with soft bokeh, pale blue sky, blurred buildings, and warm summer light; frame the scene with a few pale pink and white flowers along the right edge. Use a polished Japanese anime illustration style with delicate linework, luminous pastel coloring, soft rim light, airy highlights, flowing hair motion, shallow depth of field, and a fresh morning atmosphere. Vertical portrait composition, 2:3 aspect ratio, no text, no watermark, no extra characters.

Create a vertical dark-pop horror anime streetwear illustration set in a narrow urban alley covered wall-to-wall with vivid graffiti. Two young punk girls stand full-body in the center, slightly low-angle perspective, facing the viewer with cool, defiant expressions. The scene should feel like Harajuku/Shibuya alternative fashion mixed with cute horror: neon pink, cyan, yellow, black, and white paint splashes, skulls, crosses, cartoon ghosts, bats, stitched monster faces, and spray-painted tags on both walls and the wet pavement. The sky above the alley is bright turquoise blue, framed by tall grimy buildings, fire escapes, pipes, and dense graffiti murals. The girls are close together and similar in height, with pale skin, dark eyes, black hair with bangs, and long side braids or loose strands. Left girl: black oversized hoodie with a large pink creepy-cute cat graphic, black beanie with a small skull/cat patch, ripped black cargo pants with pink straps, dangling chains, star and cross patches, plush skull charms, and chunky black-and-white platform sneakers with hot pink laces. Right girl: white oversized hoodie with a black-and-pink cartoon monster face graphic, pink beret or beanie with a small horror patch, ripped black cargo pants with pink cross patch, chains, charms, and matching chunky platform sneakers. Include exactly 2 main characters. Include many graffiti motifs but make the most visible recurring motifs countable: 5 large cute-horror creature faces on the walls, 3 large pink crosses on clothing or wall art, 4 small skull charms or skull icons on outfits, and 2 large blue/yellow graffiti word shapes on the right wall. Use detailed Japanese anime linework, gritty ink texture, halftone dots, comic-book shading, high contrast, saturated neon colors, paint splatter, distressed fabric, and cinematic low-angle composition. Avoid readable brand logos; any lettering should be stylized graffiti-like gibberish. Theme: {argument name="theme" default="dark pop horror"}; color palette: {argument name="color palette" default="neon pink, cyan, yellow, black, and white"}; setting: {argument name="setting" default="narrow graffiti-covered urban alley"}; character style: {argument name="character style" default="cute punk anime girls"}; outfit focus: {argument name="outfit focus" default="oversized hoodies, ripped cargo pants, chains, charms, and platform sneakers"}.

Create a cinematic photorealistic fantasy portrait of a proud anthropomorphic hamster warrior general standing on a rocky cliff at sunset. The hamster is chubby and golden-brown with a cream belly, realistic soft fur, tiny rounded ears, long whiskers, small dark confident eyes, and a smug satisfied smile. It stands upright in a heroic low-angle composition with both small arms crossed over its chest, looking slightly upward and to the right as if surveying a battlefield. Dress it in ornate ancient warlord armor: two dark bronze shoulder pauldrons with gold trim and swirling engraved patterns, layered leather/metal chest details, a red scarf at the neck, and a dark crimson cape hanging behind the left side. The setting is an epic mountain overlook with rugged rocks in the foreground and distant hazy mountains. The sky is filled with dramatic orange-gold sunset clouds, with the sun low near the right horizon creating warm rim light around the hamster’s fur and armor. Use ultra-detailed CGI realism, cinematic lighting, shallow depth of field, warm color grading, heroic fantasy movie atmosphere, no text, no watermark. Optional subject customization: {argument name="animal warrior" default="anthropomorphic hamster warrior general"}, {argument name="armor style" default="ornate ancient bronze warlord armor"}, {argument name="cape color" default="dark crimson"}, {argument name="pose" default="arms crossed with a smug proud smile"}, {argument name="background" default="rocky mountain cliff at golden sunset"}.

Create a 16:9 promotional screenshot collage for a retro puzzle-action game called {argument name="game title" default="ConsoleChaos"}, showing the same side-scrolling rooftop stage rendered in four different console-generation styles at once. Divide the canvas into an exact 2x2 grid with thin black borders between panels, like a technical pilot comparison screen. Use strong CRT scanlines across the entire image, low-resolution pixel-art aesthetics, slightly blurry capture artifacts, and dark cyberpunk city colors.

Top-left panel: 8-bit / Famicom-style side-view platformer. Magenta-purple night sky, flat pixelated buildings, yellow-green tiled ground, three simple brown vertical towers with small pale windows. A tiny pink-haired anime girl sprite in white/pink is mid-jump moving right. Very limited palette, chunky pixels, high contrast.

Top-right panel: 16-bit / Super Famicom-style side-view platformer. Bright blue sky with long flat white clouds, green rooftop fence/platform, blue building facade with rows of yellow-lit windows. The same pink-haired anime girl sprite stands on the platform facing right, wearing a blue outfit. Cleaner sprites and more colors than the top-left panel.

Bottom-left panel: early PlayStation-style 3D view. Third-person camera behind and slightly to the side of a low-poly pink-haired anime girl character standing on a gray tiled rooftop. Purple foggy night skyline, blocky buildings with a few lit windows, visible polygon edges and jagged texture warping. Add a black UI strip along the bottom with tiny white Japanese debug-like text and button hints.

Bottom-right panel: PlayStation 2-style 3D view. Third-person 3D rooftop scene at night with a larger, smoother anime girl character facing the camera, pink twin-tail hair, dark blue/black outfit with blue skirt and boots. She stands on a broad gray tiled rooftop with city buildings and yellow-lit windows below, a raised platform and blocky obstacle behind her, plus a small red object on the platform. Dark navy sky, more detailed geometry than bottom-left but still retro game-like. Include faint tiny white UI/debug text near the lower left edge.

Character details: use one recurring heroine, {argument name="character name" default="pink-haired anime girl"}, with pastel pink hair and a blue/black futuristic school-uniform-inspired outfit in the later-generation panels. Keep the four render styles clearly distinct: 8-bit, 16-bit, PS1 low-poly, and PS2-era 3D. Do not add a title logo or extra explanatory captions; this should look like a captured development comparison image rather than a polished poster.

Create a vibrant square manga-style festival illustration of {argument name="animal performers" default="cute hamsters"} performing Japanese taiko drums at night during a matsuri. The composition is packed and energetic, with three large foreground hamster drummers as the focus: the center hamster is largest, wearing a red happi coat and blue-and-white twisted headband, smiling with mouth open and holding two drumsticks high over a large brown taiko; the left hamster wears a blue happi coat, winks, and plays a matching taiko; the right hamster wears a dark navy happi coat, smiles with eyes closed, and plays another matching taiko. Include exactly five visible foreground taiko drums: three main full drums across the front and two partial cropped drums at the far left and far right edges. Behind them, show a dense cheering crowd of smaller hamster drummers in festival headbands and red/blue coats, holding drumsticks, with lantern strings and festival banners reading 「祭」 on both sides. Use a dramatic golden radial burst background with black speed lines converging upward, falling multicolored confetti, small star shapes, and warm orange festival lighting. Add large bold manga sound effects in Japanese, exactly five visible instances of 「ドン」 in red-to-black gradient letters with thick white outlines, arranged across the upper half. In the top-right corner, place a white rectangular narration box with black border containing vertical Japanese text: {argument name="narration box text" default="その宴は三日三晩続いた"}. Style should be highly polished cute-anime illustration, glossy eyes, rosy cheeks, thick ink outlines, exaggerated joy, detailed drum ropes and wood texture, dynamic comic-book motion, saturated colors, no photorealism, no watermark.

Goal: Create a bold experimental product poster for a design-object announcement, combining an extreme cropped typography background with a thermal receipt layout overlay.

Canvas: Vertical poster, 4:5 aspect ratio, off-white outer margin around a saturated hot-pink rectangular field. Use a clean print-design composition with slight risograph/thermal-paper grain and subtly imperfect ink edges.

Layout: Place one huge cream-white uppercase background title reading {argument name="product name" default="TABLE LAMP"}, spanning nearly the full width across the upper third; crop the letters heavily at the left and right edges so the title feels oversized and partially hidden. Center-right, overlay exactly one tall narrow cream thermal receipt running from near the top to near the bottom, with zigzag torn edges at the top and bottom. Add one thin vertical microtext line beside the receipt, rotated 90 degrees, reading like an archive/design ID. Along the bottom of the pink field place exactly two separate cream barcode blocks, one at bottom left and one at bottom right. The receipt itself contains exactly one black barcode near its bottom.

Receipt details: The receipt header has a large condensed black title {argument name="receipt header" default="TABLE LAMP"}. Under it, include the subtitle “AESTHETICS, SWITCHED ON”. Fill the receipt with monospaced thermal-printer text, dashed divider lines, and pseudo-transaction metadata. Include design-spec lines such as design registration number, date, time, terminal, transaction, attendant, product, object type, form study, material study, surface finish, light quality, curve ratio, form revision, assembly check, aesthetic value, archive index, subtotal, tax, payment method, authorization, thank-you note, return policy, and “KEEP THIS LIGHT RECEIPT”. Highlight exactly one line in a translucent pink marker strip: “GRAND TOTAL--BEAUTY, SWITCHED ON”.

Subject image on receipt: At the lower half of the receipt, print exactly one cute glossy flower-shaped table lamp illustration with a pink fluted pedestal stem, six rounded pink petals, a blue inner glow, a small orange-red circular center, and a thin pale power cord trailing left. Keep it centered above the receipt barcode.

Typography and style: Use condensed bold sans-serif for the massive background title, monospaced receipt type for all receipt text, black ink on warm cream paper, and a fashionable editorial graphic-design aesthetic. Main colors are {argument name="background color" default="hot pink"}, cream white, black, soft pink, blue, and orange-red. The image should feel like a modern poster generated from the structural idea of a thermal-paper receipt, not a realistic store receipt photo.

Constraints: Use exactly 3 visible barcodes total: 2 large background barcode blocks and 1 receipt barcode. Use exactly 1 receipt, exactly 1 product illustration, and exactly 1 highlighted grand-total line. Do not add people, logos, watermarks, QR codes, extra products, or extra receipts.

Goal: Create a vertical celebratory 4-panel anime manga illustration about a cute moon-themed princess reacting to a video view count milestone.

Canvas: Tall vertical webcomic format, about 9:16 aspect ratio, with four stacked rectangular panels separated by thick dark borders. Use bright, polished gpt-image-2 style anime rendering, clean line art, soft gradients, expressive chibi proportions, and energetic manga effects.

Main character: A small chibi anime princess with short bobbed {argument name="hair color" default="blue-purple"} hair, large sparkling blue eyes with star highlights, rosy cheeks, and a gold tiara topped with a blue heart gem. She wears a royal purple-blue winter princess outfit: cape/robe with gold trim and star patterns, fluffy white fur collar and cuffs, decorative bow at the chest, layered white skirt, black thigh-high socks, and blue shoes. Keep her design consistent across all four panels.

Panel count and layout: Use exactly 4 panels.
1. Top panel: Cozy room at night with a window showing a red moon in a starry sky, purple curtains, snowflake decorations, a mug on the desk, a smartphone in the character’s hand, and a purple book/card on the right. The character looks curious and thoughtful while checking the phone. Add one large round speech bubble on the left containing the text {argument name="question text" default="RED MOON\n再生数\nいくつかな？"}. Include a small purple question mark near the right side.
2. Second panel: Extreme close-up of the character’s surprised face, mouth open, eyes wide, with a purple radial speed-line background. Add exactly one jagged speech bubble on the left containing {argument name="surprise text" default="え!?!?"}. Add one yellow surprise sparkle near the bubble.
3. Third panel: The character is seated or bouncing in shock, both hands raised, mouth open, small blue sweat-drop marks near her head, red/pink radial burst background. Add exactly one large jagged speech bubble occupying the left half, with huge bold red milestone text {argument name="milestone text" default="1.2万\n再生ええええ"}; make the number the largest element in the panel.
4. Bottom panel: Full-body joyful celebration pose, the character leaps forward with both arms spread wide, eyes closed in a huge smile. Background shows the large red moon outside the broken-looking panel frame, confetti, sparkles, snowflakes, paper shards, and energetic celebratory lines. Add exactly one large cloud-shaped speech bubble in the lower-left containing oversized bold red text {argument name="thank you text" default="ありがとうー！"}.

Visual style: Cute Japanese anime/manga chibi illustration, saturated purples and reds, glossy highlights, soft blush, dynamic comic motion lines, festive atmosphere, crisp outlines, high-detail clothing ornaments, playful celebratory tone.

Constraints: Preserve exactly four panels and exactly four main speech bubbles, one per panel. Do not add extra characters, logos, watermarks, or unrelated text. Keep all visible text large, legible, and integrated into manga-style bubbles.

Create a vertical 9:16 fantasy anime key visual in super-deformed chibi style, showing a four-member adventuring party rushing diagonally toward the viewer through a bright magical battlefield. The composition contains exactly 4 characters: 1) front lower center, a tiny white-haired mage girl with long pale lavender-white hair, purple eyes, a calm expression, wearing a flowing white robe with gold trim and holding a wand tipped with a glowing pink crystal cube, surrounded by violet magical arcs; 2) middle left, a black-haired schoolgirl samurai with golden-brown eyes, short navy sailor-style uniform, small armor details, and a drawn katana, lunging forward with a determined face; 3) middle right, a blonde elf girl with long wavy hair, pointed ears, blue-green eyes, teal cape, white tunic, dark leggings, and a short wand or rapier casting bright green circular magic sigils; 4) upper right, a stocky red-haired dwarf warrior with spiky topknot, enormous red eyebrows and beard, fierce eyes, dark red samurai armor, shoulder plates, and a sword at his waist, leaning forward protectively. Use {argument name="party size" default="four"} chibi fantasy heroes, with the central mood as {argument name="mood" default="energetic heroic adventure"}. Background: pastel watercolor splashes, star and petal confetti, sparkling particles, faint fantasy castle silhouette on the upper left, faint red torii gate on the upper right, and no readable text. Style: glossy modern Japanese anime, cute RPG party illustration, dynamic diagonal motion, soft rim lighting, vibrant rainbow magic effects, detailed costumes, clean faces with large eyes, high resolution, whimsical and celebratory. Keep the characters fully visible, overlapping but distinct, with the mage closest to the bottom foreground and the dwarf highest in the background. Avoid logos, captions, speech bubbles, borders, and photorealism.

Goal: Create a clean side-by-side comparison graphic showing two AI image model outputs as realistic street portraits.

Canvas: Horizontal 3:2 social media comparison image, white background, high resolution. Use a simple two-column layout with a narrow vertical white divider between the image panels.

Layout: Top header area contains exactly 2 black logos centered above exactly 2 bold model names. Left column: OpenAI knot logo above the heading “{argument name="left model name" default="GPT-Image 2"}”. Right column: Grok-style black circular slash logo above the heading “{argument name="right model name" default="Imagine Image 2.0"}”. Below the headers are exactly 2 equal-size rectangular photo panels: one on the left and one on the right.

Photo content: Both panels show a cinematic, photorealistic urban street portrait of the same young woman in a crowded city. She has {argument name="hair color" default="dark brown"} hair, natural makeup, full brows, soft lips, and a calm direct gaze. She wears a dark coat. The background is a busy city sidewalk with motion-blurred pedestrians, shallow depth of field, muted gray urban tones, and warm bokeh traffic lights.

Left photo panel: The woman is slightly left of center, turned partly toward the camera, with loose voluminous hair around her face. A large blurred pedestrian crosses the foreground on the right, partially obscuring the lower face/body area; another blurred figure appears on the left. The image should feel like a candid real-camera street photograph with natural imperfections and documentary realism.

Right photo panel: The woman is centered closer to camera, facing forward with slightly tousled hair pulled back loosely and face-framing strands. Her facial detail is sharper and more polished, with visible skin texture and freckles. Blurred pedestrians pass on both sides, including a gray figure on the left and a warm brown motion blur on the right.

Visual style: Photorealistic DSLR street photography, natural light, shallow depth of field, 50mm lens feel, realistic skin texture, subtle filmic color grading, high detail in faces, strong motion blur on passing pedestrians, no illustration style.

Text and element count: Include exactly 2 logos, exactly 2 headings, exactly 2 photo panels, and exactly 2 main female portraits. Do not add extra captions, scores, labels, borders, watermarks, or interface elements.

Create an ultra-realistic cinematic close-up studio portrait of a {argument name="subject" default="confident young adult Black man"} facing the camera directly, centered in the frame against a completely dark black background. He has a short neat fade haircut, smooth skin with natural pores and subtle highlights, dark focused eyes, relaxed closed lips, and a calm powerful expression. Pose him with one hand thoughtfully under his chin, thumb along the jawline, showing a polished silver metal wristwatch with a black dial on the wrist. Dress him in a plain {argument name="shirt color" default="black"} crew-neck T-shirt. Use dramatic split lighting: saturated electric {argument name="left light color" default="blue"} light illuminating the viewer-left side of his face, hand, and shoulder, and saturated {argument name="right light color" default="red"} light illuminating the viewer-right side, with a subtle blend across the center of the face. Make the image feel premium, masculine, mysterious, editorial, and powerful, like a high-end magazine cover portrait. Use tight head-and-shoulders framing, shallow depth of field, crisp facial detail, realistic skin texture, glossy catchlights in the eyes, strong contrast, deep shadows, no visible text, no logos, no watermark, no props besides the watch, and no background detail.

Create a photorealistic cinematic fashion portrait of two adult East Asian women standing close together in a cozy bedroom at golden hour, both turned away from the camera but looking back over their shoulders with calm, serious expressions. The woman on the left wears a fitted ribbed slip dress in {argument name="left dress color" default="pale powder blue"} with thin straps, long slightly tousled dark hair, soft natural makeup, and dewy skin; the woman on the right wears a fitted ribbed slip dress in {argument name="right dress color" default="deep red"} with thin straps and an open back, shoulder-length dark hair, natural makeup, and warm highlights on her face and back. Their poses are elegant and intimate, with one arm gently around the other's waist, emphasizing fashion editorial styling rather than explicitness. Set the scene in a small hotel-like bedroom with exactly 6 main visible background elements: an unmade white bed with pillows on the left, a dark headboard, a framed landscape photo above the bed, tall sheer curtains on the right, a bright window behind the curtains, and a small bedside table near the window. Use a wide-angle fisheye lens look with curved distortion and dark vignette edges, as if shot on a {argument name="camera lens" default="16mm fisheye lens"}. Lighting should be warm sunset light streaming through the curtains, creating soft shadows, glowing skin highlights, and a moody amber atmosphere. Compose in a horizontal 16:9 frame, close mid-shot from behind at waist-to-shoulder height, shallow depth of field, high dynamic range, realistic skin texture, cinematic color grading, no text, no watermark, no extra people.

Goal: Create a cute digital sticker pack preview sheet for messaging apps, showing exactly 9 chibi stickers in a 3 by 3 grid on a clean white canvas. The main character is {argument name="character description" default="a white-haired anime cat girl with pale pink gradient twin-tail hair, blue eyes, white cat ears, a pink cat tail, a loose blue hoodie, black pleated skirt, yellow socks, and sneakers"}. Her sidekick is {argument name="mascot description" default="a small round navy-blue penguin wearing a matching blue hoodie, sometimes holding a tiny phone"}. Use a soft kawaii Japanese LINE-sticker style: rounded forms, pastel colors, thick clean outlines, expressive faces, small sparkles, hearts, motion lines, and hand-drawn bubble lettering. Canvas: square-ish white sticker sheet with rounded corners, no heavy background, subtle drop shadow, and small miniature sticker-sheet thumbnails stacked vertically along the right edge like a marketplace preview. Layout: exactly 9 large stickers, evenly spaced in three rows and three columns. Include these 9 sticker scenes and matching text phrases from {argument name="sticker phrases" default="おはよ〜, ありがとっ!, おつかれさま〜, OKだよ〜, ごめーん, いま行くー!, よろしくね, おやすみ〜, またね!"}: 1) top left, cheerful morning greeting: cat girl waves with one hand, penguin smiles beside her, yellow-orange handwritten text above; 2) top center, thank-you: cat girl bows with eyes closed and hands down, penguin bows too, orange text with a small heart; 3) top right, good-work/tired greeting: cat girl salutes while holding or standing near the penguin, blue text and sparkles; 4) middle left, OK confirmation: cat girl gives a big thumbs-up with a wink, penguin laughs, green text with stars; 5) middle center, apology: cat girl clasps hands apologetically with sad watery eyes, penguin bows deeply, blue text with sweat drops; 6) middle right, on-my-way action: cat girl runs energetically with flowing hair and tail, penguin runs beside her holding a phone, red-orange text and motion lines; 7) bottom left, please/nice-to-meet-you: cat girl and penguin bow politely, pink text with a small heart; 8) bottom center, good night: sleepy cat girl hugs a blue star-pattern blanket or pillow, penguin wears a blue nightcap and hoodie, purple text with crescent moon and stars; 9) bottom right, see you again: cat girl waves happily, penguin waves beside her, orange text with excitement marks. Add a small semi-transparent gray circular button labeled {argument name="edit button text" default="編集"} at the bottom-left corner and a small semi-transparent circular share/download icon at the bottom-right corner, like a mobile preview overlay. Constraints: keep the characters consistent across all 9 stickers, make the text large and hand-lettered above or near each sticker, use no photorealism, no extra characters, no extra sticker panels, and no watermark.

Goal: Create a high-end cinematic luxury skincare advertisement for {argument name="product name" default="Vitamin C Serum"}, focused on radiant skin, golden-orange serum, and fresh oranges.

Canvas: Vertical 3:4 image, split into exactly 2 horizontal panels separated by a thin white divider line. Warm golden-hour sunlight, shallow depth of field, glossy reflective surfaces, premium beauty-ad photography.

Subject details: A beautiful young woman with {argument name="hair color" default="long dark brown hair"}, natural glowing skin, soft elegant facial features, thick brows, minimal warm makeup, small stud earrings, a delicate necklace, and a gentle confident smile. She wears a simple white camisole. Her expression is calm, fresh, and natural, looking toward the camera.

Top panel: Close-up beauty shot of the woman holding a glass dropper with a metallic gold cap near her cheek. A single golden-orange serum droplet is visibly hanging from the pipette and gliding onto her skin. Place one amber glass skincare bottle on the right foreground with a gold label reading “VITAMIN C SERUM” and smaller text “BRIGHTENING & GLOW BOOSTING” plus “30ml e 1.01 fl.oz.” Include exactly 2 visible orange elements in this panel: 1 large cut orange half in the lower-left foreground and 1 whole/partial orange behind the bottle on the right. Background: softly blurred sunlit window, beige curtains, indoor greenery.

Bottom panel: Show the same woman gently massaging serum into her right cheek with her fingertips, maintaining a soft confident smile and radiant hydrated skin. Place one amber dropper bottle on a small clear glass pedestal at the left foreground, with the white dropper top visible and the same gold “VITAMIN C SERUM” label. Include exactly 5 visible orange elements in this panel: 1 whole orange on the far left, 1 cut orange half/slice at the lower left, 1 orange wedge near the center foreground, 1 large cut orange slice in the lower-right foreground, and 1 whole/partial orange on the far right. Add a few green leaves around the oranges and warm bokeh highlights in the background.

Visual style: Ultra-realistic commercial beauty photography, luxurious amber and citrus color palette, warm sunlight, natural skin texture, dewy highlights, realistic serum gloss, polished reflections, soft cinematic focus, premium skincare advertising composition.

Constraints: Use exactly 2 stacked panels, exactly 2 product bottle appearances, exactly 1 visible serum droplet in the top panel, and no extra text beyond the product label.

Create a cinematic anime short-film still of a dynamic jewelry heist fight inside a luxurious modern shopping mall. The frame is a horizontal 16:9 shot with a slightly wide-angle camera, shallow depth of field, subtle motion blur, and bright afternoon sunlight streaming through a huge glass skylight ceiling. In the foreground, show a sleek illuminated glass jewelry display case filled with white velvet trays, gold necklaces, rings, and sparkling diamonds; the main character is an athletic anime heroine, {argument name="character name" default="Rin"}, wearing a black cyber-ninja streetwear outfit with cropped top, arm straps, short skirt panels, thigh straps, and chunky black sneakers with pink accents. She is mid-leap on top of the glass case, one knee raised, long flowing {argument name="hair color" default="black"} hair whipping dramatically behind her, holding a small black pistol or tactical gadget while twisting away from an attacker. Include exactly 1 masked robber in the left foreground, dressed head-to-toe in matte black tactical clothing and balaclava, crouching and lunging toward her. Background details: a high-end mall atrium with polished floors, curved balconies, warm ceiling lights, glass railings, a tiered crystal perfume display tower at center, elegant jewelry counters, and one visible luxury storefront sign reading {argument name="storefront sign" default="DANCE"}. Mood is fast, glamorous, and tense, like a 15-second anime action clip frozen at the most exciting moment. Color palette: warm champagne mall lighting, creamy whites, reflective glass, black clothing, pink shoe accents, and sparkling highlights. Add a video-player style presentation around the image: black dotted interface background, rounded-rectangle preview frame with thin gray border, small camera icon and title text at top left reading {argument name="video title" default="Anime Mall Jewelry Heist Fight"}, a small expand icon in the top-right corner of the preview, a circular right-arrow button outside the preview on the right, and a thin progress scrubber with a white handle on the left edge. Keep all text crisp and minimal, no extra logos or watermarks.

Goal: Create a cute LINE sticker sheet featuring {argument name="character name" default="a chibi cat-eared girl"} and a small penguin mascot, arranged as nine separate transparent-style stickers on a clean white background.

Canvas: Square 1:1 image, high resolution, white background, sticker-pack preview layout with a 3 by 3 grid. Each sticker has soft pastel colors, thick clean outlines, subtle white sticker edging, tiny sparkles or hearts, and rounded kawaii proportions.

Character details: The main character is a small chibi anime girl with fluffy long {argument name="hair color" default="white hair with pale pink tips"}, cat ears, large blue eyes, a cozy oversized {argument name="outfit" default="blue hoodie"}, short skirt or shorts, and small shoes. She appears with a round dark-blue penguin mascot with a white face and yellow beak/feet. Expressions are cheerful, shy, apologetic, excited, sleepy, and affectionate.

Sticker count and labels: Include exactly 9 discrete stickers, each with handwritten Japanese text in bright pastel colors above or near the characters:
1. Top left: Girl waving happily beside penguin, text 「おはよ〜」 in yellow-orange.
2. Top center: Girl bowing or nodding with penguin at her side, text 「ありがとう！」 in orange.
3. Top right: Girl hugging or holding penguin close, text 「なかよくしよ」 in blue.
4. Middle left: Girl making an OK gesture with penguin, text 「OKだよ〜」 in green.
5. Middle center: Girl looking sleepy or relaxed with penguin, text 「ごめん〜」 in blue.
6. Middle right: Girl and penguin running or moving energetically, text 「いま行く〜！」 in red-orange.
7. Bottom left: Girl bowing politely with penguin nearby, text 「よろしくね」 in pink.
8. Bottom center: Girl lying or sitting cozily under a blue blanket with penguin, text 「おやすみ〜」 in purple.
9. Bottom right: Girl jumping excitedly with penguin, text 「またね！」 in orange.

Visual style: Polished Japanese kawaii LINE sticker style, soft watercolor-like shading, pastel blue and pink palette, expressive faces, playful hand-lettering, compact sticker compositions, no background scenery.

Constraints: Keep all nine stickers clearly separated and evenly spaced in a 3x3 grid. Preserve the Japanese text exactly as written. Do not add extra stickers, captions, logos, watermark, borders, or UI elements.

Create a vertical triptych (three tall panels side by side) in a classic {argument name="art style" default="Japanese ukiyo-e and ink-wash style"}.  
Subject: {argument name="characters" default="A young samurai girl and an older samurai man"} fighting with katanas in front of {argument name="architecture" default="old-style Japanese architecture (wooden temples, tiled roofs, shoji screens, stone lanterns)"}.  
Color and mood: Overall brownish, sepia-toned palette with an aged-paper texture. The background should be darker than the foreground, using deep browns and near-black shadows to create depth and drama.  
Composition:  
- Left panel: The two warriors facing off, swords drawn, tense stance.  
- Middle panel: Mid-clash, blades crossing, sparks and motion lines, dynamic poses.  
- Right panel: A decisive strike moment, one warrior lunging forward, the other parrying.  
Keep the characters’ faces, armor, and clothing consistent across all three panels. Use strong ink outlines, subtle watercolor washes, visible brush strokes, and muted earth tones (ochre, umber, sepia, dark brown). The overall feel should be historical, cinematic, and dramatic.

Recreate the reference image as a premium editorial city poster. Replace the original person with {argument name="subject" default="the person from the uploaded image"} while preserving {argument name="subject reference" default="SUBJECT"}'s identity, facial features, hairstyle, hair color, skin tone, and natural appearance. {argument name="subject reference" default="SUBJECT"} must be standing in the SAME pose, SAME body orientation, SAME framing, SAME scale, and approximately the SAME position as the reference subject.

Replace the original city with {argument name="city name" default="CITY_NAME"} and rebuild the surrounding skyline, landmarks, architecture, bridge, and urban photography specifically for {argument name="city name" default="CITY_NAME"}. Replace all city-specific labels, coordinates, location information, and captions accordingly.

CRITICAL TYPOGRAPHY COMPOSITION: The oversized {argument name="city name" default="CITY_NAME"} typography must be physically layered OVER the standing character, exactly like the reference. Do NOT place all of the city text behind the subject. The large {argument name="city name" default="CITY_NAME"} letters must pass across the subject's body and overlap her clothing and lower torso, visibly covering portions of her. At the same time, some portions of the typography should remain behind her head and shoulders. Create a true foreground/background collage: BACKGROUND CITY IMAGERY → SUBJECT → FOREGROUND OVERSIZED {argument name="city name" default="CITY_NAME"} LETTERS. The foreground letters must visibly obscure portions of the subject while the subject remains recognizable through the gaps and around the typography. The large letters should intersect the character naturally rather than appearing as text floating beside her.

The oversized {argument name="city name" default="CITY_NAME"} typography should dominate the composition and extend almost across the entire poster width. Integrate grayscale city photography inside selected portions of the letters. Keep the typography bold, geometric, heavy, and editorial. Make the letters large enough to cross directly over the subject's torso and lower body.

3D render of {argument name="subject" default="[SUBJECT]"}, stylized hand-painted 3D game asset, chunky simplified geometry, broad readable shapes, softly beveled edges, matte painted wood and metal, subtle hand-painted surface variation, restrained material detail, deep plum and burgundy wood, dark teal and navy hardware, selective crimson and muted warm accents, soft diffuse lighting, gentle ambient occlusion, simple soft directional shadow, flat solid black background, no vignette, no dramatic lighting, no volumetric effects, elevated isometric three-quarter view, clean centered composition, compact game-ready silhouette, understated handcrafted atmosphere.

Use my uploaded reference image as the primary facial identity reference. Preserve my exact facial identity with maximum accuracy. Keep my facial structure, face shape, jawline, cheekbones, eyebrows, eye shape, nose, lips, ears, skin tone, hairstyle, hair texture, beard/stubble pattern, and all distinctive facial characteristics recognizable and consistent. Do not beautify, feminize, de-age, or alter my identity. Create a cinematic, ultra-realistic black-and-white portrait of me sitting at a dark wooden table in a sophisticated, dimly lit interior. I am wearing a {argument name="clothing" default="loose black crew-neck T-shirt"}. I am seated naturally, leaning slightly forward, with one hand gently resting beneath my chin in a thoughtful, confident pose. My other hand rests on the table while holding a {argument name="object" default="plain white ceramic coffee cup"}. My expression is calm, serious, and introspective, with natural direct eye contact toward the camera. Use dramatic {argument name="lighting" default="side window lighting"}, illuminating one side of my face while the other falls into deep cinematic shadow. High-contrast monochrome photography, rich blacks, soft highlights, realistic skin texture, detailed beard and individual hair strands, subtle film grain, and premium editorial fashion photography. The background should be a dark modern interior with a large window or glass door, softly blurred outdoor foliage and architectural details. Strong shallow depth of field and creamy natural bokeh. Add subtle out-of-focus foliage in the foreground. Composition: vertical 4:5 portrait, upper-body framing, subject slightly off-center, coffee cup clearly visible in the lower portion of the frame. Natural proportions and realistic anatomy. Absolutely no text, no typography, no logos, no captions, no watermark, and no writing anywhere in the image. Ultra-realistic, photorealistic, cinematic monochrome, 85mm lens look, f/1.8, shallow depth of field, sophisticated, moody, timeless, high-end editorial portrait photography.