GPT Image 2 prompts

Browse GPT Image 2 prompts for fashion portraits and graphic compositions. Compare how each prompt describes its subject and the way the image is arranged.

探索提示词

9,951 条提示词

E-commerce Main Image - Pale Silver-Gray Corset and Maxi Skirt — output

E-commerce Main Image - Pale Silver-Gray Corset and Maxi Skirt

MoodLock|AI画像の惜しいを直すGPT Image 2
PROTECTED WARDROBE RULE — HIGHEST PRIORITY: Treat each garment description below as its immutable final product identity. Preserve construction, parts, hardware, material, pattern, and every registered graphic or marking—its surface, side, orientation, scale, colors, content, spelling, and count. FINAL WORN STATE overrides only the scoped item's use, position, side, orientation, fastening, layering, folds, tucks, knots, and drape. A SOURCE GARMENT label describes the unstyled item, never the completed silhouette. Do not redesign, add, remove, mirror, duplicate, relocate, or redraw protected details. Wardrobe: Top: One pale silver-gray and white strapless overbust corset top made from medium-weight structured woven corsetry fabric with a fine diagonal surface, low sheen, firm shape retention, and slim vertical boning channels. The close longline silhouette ends at the upper hip, narrowing through the waist before opening slightly over the lower side panels. A pronounced sweetheart upper edge shapes two curved bust cups around a shallow center notch, with scalloped white floral lace and fine eyelash fringe tracing the complete upper perimeter. Curved cup seams, princess seams, and narrow boning channels divide the front, side, and rear into tapered panels; a white bound edge follows the complete lower contour. White lace forms an all-over botanical field of large blossoms, small rosettes, leafy stems, curling vines, and open net regions across the pale front, both sides, and rear panels, with denser scalloped motifs along the bust edge. A vertical champagne gold corset busk closes the center front. Two white center-rear plackets carry parallel rows of antique-gold-tone grommets, crossed by one wide white satin ribbon from the straight rear upper edge to the lower edge and finished in a secure tied closure. Keep the lace edges crisp against the pale base, the boning channels narrow and straight, the busk and grommets firmly seated, and the ribbon broad and flat through every crossing. Bottom: One off-white ankle-length tiered maxi skirt made from lightweight semi-sheer floral voile, white Swiss-dot dobby voile, and broad white floral mesh lace. It has a full softly flared A-line silhouette with a narrow upper contour and a wide flowing lower sweep. A narrow waistband forms the level upper opening. The main upper panel extends from the waistband to the lower calf and joins one broad gathered Swiss-dot lower tier through a level horizontal seam. Tiny coral-pink floral sprigs with faint pale-sage foliage scatter evenly across the upper panel; narrow scalloped white trim follows the tier join, and tiny raised dots cover the lower tier. A wide open floral-lace border overlays the lower portion of the dotted tier, with a wavy scalloped upper boundary, linked blossoms and leaves across fine mesh, and a deep scalloped outer hem. Concealed elastic within the waistband provides a clean pull-on closure. The rear repeats the floral field, the level gathered join, the dotted tier, and the overlaid lace border around the full circumference. Keep the voile airy and softly draped, the printed sprigs delicate and distinct, the raised dots regular, and the lace open and finely resolved. Footwear: One matching pair of champagne-gold open-toe stiletto sandals, identical except for anatomical mirroring. Each shoe has a slim rounded forefoot strap, a second fine diagonal strap crossing the instep, and a narrow slingback circuit completed by a concealed elastic join at the rear. The upper straps are smooth low-luster leather-like material with fine folded edges and minimal seam bulk. A thin champagne insole follows a narrow tapered footbed above a very shallow outsole, ending in a slender medium-height stiletto heel with a straight shaft and a compact heel tip. Keep both shoes equal in strap width, strap route, toe opening, heel height, color, section thickness, and finish. Necklace: One short delicate pale-gold necklace made from a very fine oval-link chain with an even slender section and soft warm polish. The continuous circuit consists of repeating small links, carries no pendant or station ornament, and closes with one compact spring-ring clasp and a matching terminal ring. Keep the link joins crisp enough to remain visible without increasing the chain's fine scale. Earrings: One matching pair of small polished gold-tone sculptural post-drop earrings. Each member combines a short slim top stem with one thick rounded metal bar that curves into an irregular elongated teardrop outline, narrows toward the top, and curls inward at the lower interior before returning upward, leaving a clearly open center. The pair has identical contour, scale, section thickness, and reflective finish. Low-profile posts and compact backs complete both members. Keep the rounded bar continuous, the inward curls crisp, the open centers unobstructed, and the polished reflections controlled along the curved sections. Bracelet: One broad champagne gold beaded bracelet constructed as a single flexible multi-row circuit from several closely packed parallel strands. The strands alternate large polished champagne gold round beads, smaller pale round beads, and clear faceted beads, interlaced into one dense band with controlled spacing and a softly articulated section. Concealed elastic cords close the circuit without exposed hardware. Keep the bead-size hierarchy distinct, the rows parallel through the full circumference, and the reflective metal, pale beads, and clear facets cleanly separated. Final worn state: VIEW-CONDITIONAL BODY-SIDE MAP: front view — wearer-left is image-right and wearer-right is image-left; back view — wearer-left is image-left and wearer-right is image-right. In profiles, crossed limbs, or ambiguity, trace the named anatomical side continuously from its corresponding shoulder, hip, eye, or ear landmark as applicable. Apply the state only to the named anatomical side; never mirror, swap, or duplicate it. - Bracelet placement: Render one complete instance of the registered bracelet from the item description immediately above around only the wearer-left wrist. Keep the wearer-right wrist free of this bracelet; do not mirror or duplicate it. Preserve the registered design and count. Keep other separately selected wrist items unchanged.
使用此提示词 ↗
Product Marketing - Cinematic Motorcycle Night Photography — output

Product Marketing - Cinematic Motorcycle Night Photography

HeisenLegacyGPT Image 2
Ultra-realistic cinematic night photograph of a {argument name="subject" default="strikingly handsome young man in his early 20s"} sitting casually on a {argument name="motorcycle" default="large black classic motorcycle"} outside a vintage motorcycle shop at night. He has a naturally attractive, masculine face, thick slightly wavy jet-black medium-length hair falling naturally around his forehead and ears, strong well-defined eyebrows, deep expressive dark-brown eyes, a straight sculpted nose, prominent cheekbones, subtle stubble and a defined jawline. He wears small black vintage sunglasses, a {argument name="clothing" default="loose oversized off-white T-shirt"} with casually rolled sleeves, relaxed black trousers, a simple thin necklace and a dark wristwatch. His forearms have tasteful realistic tattoos.
He sits in a relaxed, confident pose with one arm resting naturally on the motorcycle, looking slightly away from the camera with a calm, mysterious expression. Large glossy black vintage cruiser motorcycle dominates the foreground, with detailed chrome engine components, handlebars, mirrors, fuel tank and realistic reflections.
Behind him is a small old-school motorcycle shop, its windows completely covered with vintage motorcycle posters, photographs and advertisements. A glowing neon shop sign is partially visible above the storefront. Warm orange/red neon lighting mixes with cool cyan-green fluorescent light spilling from inside the shop, creating a dramatic teal-and-orange cinematic color contrast.
Shot at night, moody urban atmosphere, deep shadows, realistic skin texture, natural facial details, subtle film grain, shallow depth of field, beautiful bokeh from street lights, atmospheric reflections on chrome, slightly underexposed cinematic lighting, authentic candid photography, 35mm film aesthetic, high dynamic range, photorealistic, editorial motorcycle photography, dramatic composition, low-angle perspective, highly detailed, 8K realism.
Important: completely original fictional person, natural realistic face, no resemblance to any real person, no distorted hands or facial features, no artificial/plastic skin, no text or watermark.
使用此提示词 ↗
Comic / Storyboard - Personalized Rider-Waite Tarot Card — output
Profile / Avatar - Black and White Pencil Sketch Portrait — output

Profile / Avatar - Charcoal Sketch Portrait Illustration

lovimg_comGPT Image 2
Use the uploaded photo as the only identity reference for this person.

Convert this person into a {argument name="style" default="high-end black and white artistic portrait"}: sitting comfortably, {argument name="activity" default="reading an open book"} while holding a {argument name="object" default="small coffee cup"} in one hand.

Preserve identity features:
Maintain the person's recognizable facial structure, eyes, eyebrows, nose, lips, jawline, skin features, hairstyle, hair texture, and overall likeness, faithful to the uploaded reference photo.

Pose:
The person sits in a relaxed, elegant posture, naturally looking down at the open book while reading. One hand gently holds a ceramic coffee cup, and the other hand supports or holds the open book. The pose should be natural, quiet, and unposed.

Clothing:
Retain the person's clothing from the reference photo where appropriate, showing realistic fabric textures and natural folds. The overall outfit should be sophisticated and understated.

Artistic Style:
Transform the entire scene into a fine monochrome pencil and charcoal sketch, combined with a realistic editorial portrait photography style.

Use extremely fine graphite pencil lines, delicate cross-hatching, realistic charcoal shading, fine paper texture, subtle ink strokes, hand-drawn edges, and soft tonal transitions.

Smoke Effect:
The person is surrounded by gracefully flowing charcoal smoke, graphite dust, fine sketch particles, and ink lines. The smoke should naturally emerge around the person's hair, shoulders, arms, clothing, book, and lower body, gradually fading into the white background.

Coffee Cup:
The ceramic coffee cup must be clearly identifiable in the person's hand, rendered entirely with realistic pencil and charcoal shading.

Book:
The open book must be clearly visible and detailed, with realistic pages and a subtle handwritten-like page texture. Do not add readable text or logos to the pages.

Background:
Clean white or warm white textured paper background. Minimalist composition, with the person emerging from the surrounding smoky sketch art.

Color:
Strictly black and white and soft gray monochrome. No color at all.

Lighting:
Soft cinematic portrait lighting translated into monochrome tones, with rich highlight details on the face and natural shadows.

Quality:
Ultra-fine, realistic body structure, highly detailed hair strands, realistic hands and fingers, detailed clothing, realistic ceramic cup and book, professional gallery-grade pencil and charcoal sketch.

Composition:
Centered portrait, showing the person, coffee cup, and book completely. The head, hair, hands, cup, book, and body are all within the frame. Leave plenty of clean white space around the edges of the smoky art.

Important:
The uploaded image is for identity reference only.
Preserve the person's identity and likeness.
Do not replace the face.
Do not add text, logos, watermarks, borders, or picture-in-picture frames.
Do not make a collage.
Do not make it in color.
Do not make it cartoonish or 3D.
Do not crop the coffee cup or book.
The final result should look like a realistic photo transformed into a sophisticated black and white pencil and charcoal masterpiece.

Aspect Ratio: 4:5
使用此提示词 ↗
Product Marketing - Peacock-Inspired Modern Fashion — output
Product Marketing - Elegant Abstract Ribbon Illustration Prompt — output
Profile / Avatar - Identity-Preserving Romantic Couple Portrait — output
Product Marketing - Ivory Bloom Luxury Fashion Poster — output
Product Marketing - Surrealist Nike Sportswear Poster — output
Product Marketing - Minimalist 3D Miniature Coffee Shop — output
Comic / Storyboard - Cinematic Character Bible Sheet — output
Profile / Avatar - Realistic East Asian Woman Portrait — output

Profile / Avatar - Realistic East Asian Woman Portrait

FeyberGPT Image 2
A highly realistic close-up portrait of a {argument name="subject" default="young East Asian woman (early 20s)"} with fair porcelain skin and soft natural makeup. She has long, straight, silky dark brown hair with soft curtain bangs falling over her forehead. She wears {argument name="accessory" default="thin black rectangular glasses with subtle gold metal accents"} on the hinges and bridge. Her eyes are large and expressive with soft brown irises, looking directly at the camera with a gentle, slightly playful expression. She has a soft closed-mouth smile with glossy, natural pink lips and a light flush of pink on her cheeks.

Her right hand is raised near her face in a {argument name="pose" default="classic V-sign (peace sign)"}, fingers slightly relaxed, showing delicate fingers with two thin gold rings on the ring and middle fingers. She is wearing a thin black spaghetti-strap top that reveals her bare shoulders and collarbone.

Soft, even indoor lighting creates a gentle glow on her skin with subtle highlights on her cheekbones and the frames of her glasses. Shallow depth of field, smooth skin texture, realistic hair strands, and natural facial proportions. Neutral beige-to-cream background, clean and minimal. Shot in a high-resolution, photorealistic style resembling a professional smartphone selfie or beauty portrait, slightly tilted head pose, intimate and cute aesthetic.
使用此提示词 ↗
Infographic / Edu Visual - Anime Woman Pointing at Growth Arrow — output

Infographic / Edu Visual - Anime Woman Pointing at Growth Arrow

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)GPT Image 2
Create a cinematic anime-style single-panel illustration of {argument name="character name" default="a young woman"} in a snowy open landscape under a vivid deep-blue winter sky. She is shown from the waist up on the right side of a horizontal 3:2 canvas, smiling softly and looking toward her raised hand as she points with her index finger. She has {argument name="hair color" default="short tousled dark brown hair"} blown by the wind, large gray-blue eyes, delicate facial shading, and a calm confident expression. She wears exactly 2 visible outfit layers: a chunky cream ribbed turtleneck sweater and a cobalt-blue insulated work jacket with stitched seams, metal snaps, a high collar, and small red zipper pulls. Her fingertip touches the corner of a glowing, pixelated white zigzag arrow/line chart floating in the air on the left; the chart has exactly 1 continuous jagged line ending in an upward arrow, like a retro 8-bit financial growth graph. The background shows a bright snowfield with low snow-covered hills, exactly 2 visible steel power pylons in the distance on the left, and multiple black overhead power cables running diagonally across the sky from upper left to upper right, creating strong perspective lines. Use crisp cel-shaded anime rendering with detailed hair strands, realistic winter light, high contrast shadows, fine fabric texture, clean linework, and a slightly wide-angle composition. No text, no watermark, no extra characters.
使用此提示词 ↗
Profile / Avatar - Clean Vector Portrait of Woman with Glasses — output
Product Marketing - Neon Glass Energy Orb — output
Profile / Avatar - Soft Anime Girl Portrait — output
Infographic / Edu Visual - Buddhist Mural Composition Diptych — output

Infographic / Edu Visual - Buddhist Mural Composition Diptych

楊哥 | Yang OnchainGPT Image 2
Goal: Create a wide horizontal diptych comparing two Buddhist mural composition strategies in an aged Dunhuang / Chinese temple fresco style, showing the difference between a flat repetitive figure wall and a layered sacred assembly scene.

Canvas: Extra-wide panoramic image, about 2:1 aspect ratio, split vertically into two equal panels with no modern border, both panels painted on weathered tan plaster with faded mineral pigments, cracks, patina, and antique hand-painted texture.

Left panel: Depict a flat frontal mural filled with exactly 48 nearly identical seated Buddha figures arranged in a strict grid of 6 rows and 8 columns. Each Buddha has a round green-blue halo, dark blue hair, calm face, orange-red robe, and seated meditation pose. Keep the figures similar in size and spacing so the panel feels visually crowded and wall-like. Add decorative hanging canopy fabric at the top and a narrow floral border along the bottom. The background should be shallow, with almost no spatial depth.

Right panel: Depict a layered Pure Land / Buddhist court assembly with clear hierarchy, depth, and breathing space. Place exactly 1 large central Buddha seated on a lotus pedestal beneath an ornate circular mandorla and hanging canopy, making this the first visual focus. Flank the central Buddha with exactly 2 standing bodhisattvas, one on each side, wearing jeweled crowns, green and gold garments, and halos. Add exactly 14 larger foreground and midground monks seated or kneeling in orange-brown robes, arranged in curved rows that lead the eye toward the central Buddha. Behind them, include two distant crowd groups of many smaller worshippers on the left and right, intentionally reduced in scale and detail to create depth. Include layered mountains, cloud banks, temple pavilions at the far left and far right edges, lotus flowers, rocks, and soft misty pathways receding into the background.

Visual style: Ancient Chinese Buddhist mural, inspired by Dunhuang Pure Land paintings, Chaoyuan Tu-style hierarchical crowd composition, and Song/Yuan temple fresco aesthetics. Use muted mineral colors: ochre, sienna, cinnabar red, malachite green, teal blue, faded gold, ivory, and smoky gray. Lines should be delicate but aged, with hand-brushed outlines, flattened perspective, decorative symmetry, and subtle atmospheric depth on the right panel.

Composition intent: The left side must feel repetitive, dense, and spatially flat; the right side must feel organized, ceremonial, and deep, with hierarchy, foreground-middle-background separation, and intentional empty space around the central deity.

Text: No captions, no labels, no watermark, no modern typography.

Customizable subject: The central figure may be {argument name="central deity" default="Buddha"}; the mural tradition may be {argument name="mural tradition" default="Dunhuang Pure Land Buddhist fresco"}; the left grid count must remain {argument name="left panel figure count" default="48"}; the right main assembly should keep {argument name="foreground monk count" default="14"}; use an antique palette dominated by {argument name="primary color palette" default="ochre, cinnabar red, malachite green, teal blue, faded gold"}.
使用此提示词 ↗
Comic / Storyboard - Watercolor Kimono Smoker by Window — output

Comic / Storyboard - Watercolor Kimono Smoker by Window

🐹🐹GPT Image 2
Create a delicate vertical anime-style watercolor illustration of {argument name="character name" default="a serene young man"} seated on a wooden floor beside an open shoji-style window at dusk. He has {argument name="hair color" default="misty blue-gray"} long, tousled hair braided over one shoulder, pale skin, soft half-closed eyes, and a calm introspective expression as he exhales smoke. Dress him in a loose traditional Japanese kimono: white inner robes, a deep indigo haori, and a dark sash, with flowing fabric folds pooling around his seated legs. He holds a slender kiseru-style smoking pipe horizontally in one raised hand near his mouth, with a thin ribbon of smoke drifting upward in a loose S curve. In the foreground, place exactly 1 round ceramic incense tray containing exactly 1 green spiral mosquito coil with exactly 1 glowing orange ember; a second translucent smoke ribbon rises from the coil and curls along the right side of the figure. The background should be mostly empty warm off-white paper, with exactly 1 dark wooden window frame on the right edge and a small glimpse of water, reeds, and a low orange sunset outside. Use airy ink linework, pale blue-gray washes, subtle beige shadows, visible sketch texture, soft negative space, and a quiet melancholic summer evening mood. Keep the composition tall and uncluttered, with the character occupying the lower-right center and large blank space above the smoke; no text, no watermark, no modern objects.
使用此提示词 ↗
Profile / Avatar - Polished Anime Edit from Rough Sketch — output

Profile / Avatar - Polished Anime Edit from Rough Sketch

Int.LabGPT Image 2
Using REFERENCE_0 as the rough sketch and composition guide, clean it up into a polished, high-quality Japanese animation-style illustration with an artistic, creator-driven look. Keep the same main subject, the short turquoise hair, glasses, playful expression, cropped top silhouette, and both hands making finger-gun gestures, but refine the anatomy, hands, clothing folds, face, and linework into a finished illustration rather than a sketch.

Style and rendering: Use crisp expressive anime line art, painterly cel shading, high contrast, detailed fabric texture, and vivid cyan-pink accents. Preserve some rough sketch energy through visible scribbly marks and graphic brush textures, but make the character fully resolved and attractive.

Composition changes: Reframe the pose into a more dynamic low-angle, three-quarter view with the character leaning toward the viewer. Replace the simple pastel background with a dense abstract interior/gallery-like space made of curved hanging panels, paper sheets, sketch marks, cables, and blocks of cyan and hot pink paint.

Text overlay: Add exactly 1 rounded white label near the top right containing the Japanese text {argument name="top right label" default="スタイルなし(stylize 800)"}. Do not keep the original label from the reference.

Constraints: Do not add extra characters. Avoid photorealism. Keep the image square. Make the final result look like a refined edit of the rough draft, not a completely unrelated character.
使用此提示词 ↗
Comic / Storyboard - Industrial Factory Storyboard — output

Comic / Storyboard - Industrial Factory Storyboard

タナベ | AI動画 × マーケティングGPT Image 2
Goal: Create a cinematic 4-panel storyboard for a serious business documentary scene set inside an old Japanese metalworking factory, showing craftsmanship, inspection, and trust-building between management and workers.

Canvas: 16:9 horizontal image divided into exactly 4 equal panels in a 2x2 grid, separated by clean thick white borders. No captions, no speech bubbles, no logos, no text overlays.

Visual style: Photorealistic live-action film stills, muted industrial color grading, high dynamic range, natural window light mixed with dim fluorescent factory lighting, shallow depth of field where appropriate, gritty realism, detailed metal textures, dust and oil on machinery, cinematic contrast.

Panel count and layout: Use exactly 4 storyboard panels:
1. Top-left panel: Extreme close-up of a metal lathe in operation. A spinning cylindrical metal workpiece is being cut by a tool bit, with bright curled metal shavings flying outward. Include a worker’s gloved hand and forearm near the machine for scale. Background softly blurred with factory windows and machinery.
2. Top-right panel: Wide shot down the central aisle of the factory. A young Japanese businesswoman in a navy suit and white blouse walks confidently toward the camera, centered in the aisle. An older male factory worker stands with his back turned at a machine in the midground. Old lathes, workbenches, piles of metal shavings, and large windows line both sides.
3. Bottom-left panel: Tight portrait close-up of the same young Japanese businesswoman, facing the camera directly with a calm, determined, serious expression. She has black hair neatly tied back, minimal makeup, and wears a navy blazer over a white collared shirt. The factory interior behind her is dark and blurred.
4. Bottom-right panel: Medium two-shot of the young Japanese businesswoman smiling and shaking hands with an elderly Japanese factory worker wearing a gray work jacket and dark cap. They stand among old industrial machines in the same workshop, with warm daylight from large windows behind them, conveying agreement and mutual respect.

Subject details: The main character is {argument name="main character" default="a young Japanese businesswoman"}, wearing {argument name="business outfit" default="a navy suit and white blouse"}. The factory worker is {argument name="factory worker" default="an elderly Japanese craftsman in a gray work jacket and dark cap"}. The setting is {argument name="factory setting" default="an old metalworking factory filled with lathes and industrial machines"}. The mood is {argument name="story mood" default="serious, respectful, cinematic, documentary-like"}.

Constraints: Maintain consistent character identity across all panels, use realistic anatomy and proportions, keep the white grid borders crisp, avoid illustration or anime style, avoid extra panels, avoid text, avoid futuristic or clean high-tech factory elements.
使用此提示词 ↗
Comic / Storyboard - Cartoon Polar Bear on Ice Floe — output
Infographic / Edu Visual - Kurumi Kazama Morning Board — output

Infographic / Edu Visual - Kurumi Kazama Morning Board

まゆひらaGPT Image 2
Goal: Create a bright kawaii anime morning announcement illustration for {argument name="character name" default="Kurumi Kazama"}, combining a character portrait with a classroom-style daily board about special days.

Canvas: Horizontal 16:9 image, warm sunlit room with a large window on the right showing blue sky and greenery. Soft pastel palette, clean polished anime style, cheerful school-poster composition.

Layout: Place the main character large in the center-right foreground, a wooden framed information board on the left, a rounded speech bubble at the upper right, and a cloud-shaped mini scene behind the character near the window. Add a small round table at bottom right with a vase. Include subtle decorative flowers, sparkles, and stickers.

Main character: A cute blonde anime girl with big sparkling blue eyes, light blush, gentle smile, and side ponytail fading slightly peach-pink at the ends. She wears a colorful summer yukata-style outfit with floral patterns, turquoise diagonal collar, pink sleeves, and dark navy hakama-style skirt. Her right hand is raised in a small encouraging fist. Hair accessories include exactly 4 visible items: a patterned rectangular ornament, a teal ribbon, a gold ring charm, and a small white bunny charm with dangling purple tassels. Add a matching small bunny charm on her waist bow.

Top title text: Large blue Japanese name text 「風間 くるみ」 with white outline, and below it the romanized name “Kurumi Kazama” in blue. Decorate the title with exactly 5 small motifs: two blue flowers, two pink sparkle diamonds, and one small green leafy flower sprig.

Left information board: A standing wooden easel board with the date line “{argument name="date text" default="2026年 8月31日(月曜日)"}”. The board contains exactly 3 stacked content panels:
1. Top green-outlined panel with a red ribbon label “今日は”, large text 「野菜の日」, small text 「やさいを食べて元気になろう」, and exactly 4 cute vegetable mascots: tomato, green pepper, cabbage, and carrot.
2. Middle blue-outlined panel with large text 「宿題の日」, small text 「楽しくまなんで成長しよう」, and exactly 2 school-item mascots: a smiling blue notebook labeled “NOTE” and a smiling green pencil.
3. Bottom yellow-outlined panel with a colorful pinwheel icon and text 「テーマ: カラフル風車づくり」.

Speech bubble: Rounded white speech bubble at upper right with Japanese text: “{argument name="speech bubble text" default="おはよう〜。今日の朝ごはんは、何の野菜が入ってるのかな?"}”.

Mini scene bubble: Behind the main character on the right, create a soft cloud-shaped vignette showing a chibi version of the girl outdoors with a smiling dog. It contains exactly 2 characters: the chibi girl holding one colorful pinwheel and a tan Shiba-like dog wearing a green bandana, with another colorful pinwheel beside the dog. Add sunny grass, sky, and playful motion marks.

Right table detail: A small round wooden table with a slim vase holding exactly 3 blue flowers and green leaves. Add small date/credit text near it: “2026/08/31”, “@riddi0908”, and “made by GPT Image 2”.

Visual style: High-quality modern Japanese anime illustration, crisp lines, glossy eyes, pastel gradients, soft sunlight, charming educational poster feel, sticker-like icons, cozy classroom atmosphere.

Constraints: Keep all text legible and positioned like a designed poster. Use exactly the counted elements listed above. Avoid extra panels, extra mascots, extra characters, watermarks, or photorealism.
使用此提示词 ↗
按模型、风格或主题浏览

寻找下一个方向

ASTORIE CANVAS

让灵感,继续生长。

在同一个创作空间,连接提示词、参考素材和生成结果。

打开创作画布 ↗