GPT Image 2 prompt cases

Curated GPT Image 2 prompt examples with image previews.

Browse GPT Image 2 prompt cases, compare image previews, copy reusable GPT Image 2 prompts, or open the GPT Image 2 AI image generator with a prompt prefilled.

All GPT Image 2 prompt cases

Tap a GPT Image 2 prompt case to load it into the generator, or copy the prompt and adapt it into a reusable GPT Image 2 prompt template.

Characters & Anime: Cozy Cafe Chibi Scrapbook. GPT Image 2 scene prompt preview image.
1 prompts

Cozy Cafe Chibi Scrapbook

Create a 4:5 chibi scrapbook selfie image for an uploaded clear lifestyle photo of [PERSON]. Scene fit: This scene should feel cute, personal, handmade, and clean. It is not a serious editorial poster or technical diagram. Visual direction: warm cafe diary with latte foam, receipt scraps, tiny hearts, and soft washi tape. Subject treatment: Preserve the reference person's hairstyle, outfit color, glasses, pose, and most recognizable details, then simplify them into a polished chibi mini-me sticker with rounded proportions and expressive eyes. Composition: off-center chibi, two paper scraps, one cup sticker, cream margin. Use one main chibi figure, 2-4 small diary objects, soft paper layers, neat doodles, and generous empty space. Keep the scrapbook feeling organized rather than chaotic. Color and material: cream paper, coffee brown, coral accent, soft graphite doodles. Use refined lighting, controlled contrast, and coherent material texture that matches this scene. Typography and labels: Use only a few readable diary words such as date, mood, place, or today. No long captions. Quality rules: The subject and background must form one intentional visual system. Make the prompt result match the template scene first: use ornate detail only where the scene calls for it, and use simplicity where clarity matters. Avoid realistic adult proportions, luxury campaign mood, technical callouts, heavy collage, too many stickers, unreadable handwriting, real logos, watermarks, distorted anatomy, low-resolution artifacts, cheap stock clichés, and visual noise. Set aspect ratio to 4:5.

Open template
Ads & Campaigns: Brown Sugar Milk Tea Launch Poster. GPT Image 2 scene prompt preview image.
1 prompts

Brown Sugar Milk Tea Launch Poster

Create a 9:16 GPT Image 2 beverage campaign poster for a beverage product launch for [DRINK OR BRANDLESS PRODUCT]. Scene fit: This is a brand-safe product and beverage advertising scene for original drink concepts and packaging ideas. It should feel premium, appetizing, clean, and commercially usable without using real logos or real brands. Visual direction: a rich milk tea campaign with caramel swirls, cup silhouette, tapioca rhythm, and premium dessert energy. Subject treatment: Make the drink, cup, can, bottle, ice, foam, fruit, tea leaves, bubbles, or ingredient cues the clear hero. Build the background from the beverage's flavor, texture, temperature, and material qualities so the product and layout feel like one visual system. Composition: large cup packshot centered low, geometric caramel bands, ingredient shapes, clean headline zone. Use a strong product focal point, controlled geometric shapes, clean negative space, and a readable vertical advertising hierarchy. Keep the design suitable for mobile ads, product launch posters, and social campaign assets. Color and material: warm ivory, brown sugar amber, creamy beige, glossy cup highlights. Use refined studio lighting, condensation, foam, liquid translucency, paper or glass texture, and polished product shadows where appropriate. Typography and labels: Use short readable placeholders only: product name, flavor, launch, fresh, zero sugar, cold brew, milk tea, or one concise slogan. Do not invent real brand names, certification marks, prices, or dense nutrition text. Quality rules: The image should be more specific than a generic product ad: every shape, color, and texture must support the beverage concept. Use rich detail for indulgent drinks and clean restraint for minimal wellness or tea variants. Avoid real logos, fake brands, crowded ingredient piles, unreadable microtext, messy splashes, low-resolution artifacts, cheap stock clichés, and visual noise. Set aspect ratio to 9:16.

Open template
Characters & Anime: Clean Moonlit Anime Lock Screen. GPT Image 2 scene prompt preview image.
1 prompts

Clean Moonlit Anime Lock Screen

Create a 9:16 clean anime phone wallpaper for an anime character or original character concept [SUBJECT]. Scene fit: This scene must be visually strong, clean, and usable as a phone lock screen. It should never look like a busy poster, manga page, or character collage. Visual direction: calm moonlit portrait with elegant night atmosphere. Subject treatment: Show one clear anime character or one iconic character silhouette with crisp face, readable pose, and controlled details. Keep the character instantly recognizable if a known subject is provided. Composition: character below center, large moon disk, open top space, soft foreground edge. Use a 9:16 lock-screen composition with open space near the top for time widgets, one dominant focal point, simple depth, and no clutter. Color and material: midnight blue, silver, violet shadow, smooth gradients. Use refined lighting, controlled contrast, and coherent material texture that matches this scene. Typography and labels: No readable text, no captions, no speech bubbles, no fake app UI. Abstract symbols are allowed only if they are subtle and decorative. Quality rules: The subject and background must form one intentional visual system. Make the prompt result match the template scene first: use ornate detail only where the scene calls for it, and use simplicity where clarity matters. Avoid multiple characters, crowded panels, speech bubbles, tiny UI elements, messy effects, noisy backgrounds, overfilled composition, real logos, watermarks, distorted anatomy, low-resolution artifacts, cheap stock clichés, and visual noise. Set aspect ratio to 9:16.

Open template
Creative Templates: Bedtime Forest Mini Story Grid. GPT Image 2 scene prompt preview image.
1 prompts

Bedtime Forest Mini Story Grid

Create a 4:5 complete children-friendly picture book story page for a gentle bedtime forest story about [SUBJECT]. Scene fit: This must generate a full mini story in one image, not a single illustration. The page should feel warm, safe, simple, and easy for children to follow. Story requirement: Build a complete beginning, middle, and ending across multiple small picture-book panels. The story should include: 1) character introduction, 2) a small child-safe problem or wish, 3) one or two attempts or discoveries, 4) a gentle turning point, and 5) a warm resolution. Visual direction: a warm night forest tale where a small character gets lost, follows kind clues, meets a helper, and returns home safely. Panel layout: 5 rounded story panels arranged in a clean vertical page: opening scene, small problem, helpful discovery, gentle journey, cozy ending. Every panel must be a separate small frame or rounded cell, with clear reading order from top to bottom or left to right. Each panel needs one simple action only, so the story is understandable at a glance. Character treatment: Use friendly characters with rounded shapes, kind expressions, consistent costumes or markings across all panels, and clear body language. Keep the same main character recognizable in every small frame. Composition and pacing: Use clean gutters between panels, large simple shapes, soft backgrounds, and repeated visual motifs to connect the story. Do not overcrowd any panel. Leave a small caption area inside or below each panel when text is needed. Color and material: Use warm picture-book gouache, watercolor, colored-pencil texture, cream paper grain, soft shadows, and gentle child-friendly colors that match the story mood. Typography and labels: Use 1 short readable sentence or phrase per panel at most, like a real children's picture book. Text must be large, simple, and correctly spelled. The image should still tell the story even if captions are minimal. Quality rules: The whole page must read as one coherent picture-book sequence with a complete story arc. Keep it charming, safe, and uncluttered. Avoid a single full-page scene, adult editorial style, scary faces, complex perspective, dense background detail, tiny text, speech-bubble clutter, random panels without story continuity, broken character consistency, real logos, watermarks, and low-resolution artifacts. Set aspect ratio to 4:5.

Open template
Creative Templates: Friendly Heart Cutaway Page. GPT Image 2 scene prompt preview image.
1 prompts

Friendly Heart Cutaway Page

Create a 4:5 educational cutaway picture book page for a child-friendly cutaway learning page about [SUBJECT]. Scene fit: This scene should explain inner structure clearly for children. It needs accuracy, softness, and readable labels, not scary realism. Visual direction: warm heart anatomy learning page. Subject treatment: Show the subject as a simplified section view with rounded forms, friendly scale, and accurate major parts. Avoid graphic or clinical detail. Composition: central simplified heart, clear chambers and vessels, 5 callouts. Use one large central cutaway, 4-6 callouts maximum, tidy arrows, a clear title area, and generous margins. Color and material: warm red, soft blue, cream, smooth gouache-vector hybrid. Use refined lighting, controlled contrast, and coherent material texture that matches this scene. Typography and labels: Use short readable labels only. No dense paragraphs, no tiny scientific notes. Quality rules: The subject and background must form one intentional visual system. Make the prompt result match the template scene first: use ornate detail only where the scene calls for it, and use simplicity where clarity matters. Avoid medical gore, frightening anatomy, overloaded labels, dense textbook diagrams, tiny arrows, grim clinical mood, real logos, watermarks, distorted anatomy, low-resolution artifacts, cheap stock clichés, and visual noise. Set aspect ratio to 4:5.

Open template
Portraits & Photos: Celestial Deity Editorial Portrait. GPT Image 2 scene prompt preview image.
1 prompts

Celestial Deity Editorial Portrait

Create a 9:16 luxury Eastern myth portrait poster for Xi Wangmu, the Queen Mother of the West, holding a jade ritual tablet beside peach-blossom mist and Kunlun mountain geometry. Scene fit: This scene should be ornate, sacred, poetic, and cinematic. It can be opulent, but every symbol must serve the mythic identity rather than becoming random decoration. Visual direction: regal celestial deity with museum-poster polish. Subject treatment: Portray Xi Wangmu, the Queen Mother of the West, holding a jade ritual tablet beside peach-blossom mist and Kunlun mountain geometry. First understand the figure's identity, power, temperament, story, and symbolic elements, then translate 2-4 motifs into costume, light, background geometry, and atmosphere: immortality peaches, jade authority, Kunlun terraces, ceremonial cloud halos. Keep the mythic identity recognizable without turning the image into a literal symbol pile. Composition: center or off-center hero portrait, halo architecture, layered mist. Use a strong vertical portrait, clear face and silhouette, one dominant symbolic structure, layered mist or fabric, and controlled ornamental detail. Color and material: ink black, jade, antique gold, moon white, silk texture. Use refined lighting, controlled contrast, and coherent material texture that matches this scene. Typography and labels: Use only a short title, seal-like mark, or minimal poetic label if needed. No dense lore text. Quality rules: The subject and background must form one intentional visual system. Make the prompt result match the template scene first: use ornate detail only where the scene calls for it, and use simplicity where clarity matters. Avoid cheap game splash art, random dragons, overloaded ornaments, neon fantasy clutter, fake historical paragraphs, messy symbols, real logos, watermarks, distorted anatomy, low-resolution artifacts, cheap stock clichés, and visual noise. Set aspect ratio to 9:16.

Open template
Infographics: Hangzhou West Lake Sunny City Weather Card. GPT Image 2 scene prompt preview image.
1 prompts

Hangzhou West Lake Sunny City Weather Card

Create a 4:5 luxurious city weather identity card for Hangzhou on May 18, 2026: sunny, 27°C, high 30°C, low 20°C, humidity 48%, UV 7, light southeast wind. Scene fit: This is not a plain weather app widget. It should feel like a premium city postcard, editorial travel cover, and daily weather card fused into one elegant visual system. The weather information must be readable, but the first impression should be Hangzhou's city identity. City-driven design: Use West Lake, Su Causeway, lotus leaves, distant pagoda silhouettes, Longjing tea terraces, and soft water reflections as refined motifs. Do not stack tourist icons randomly; translate them into layered borders, misty silhouettes, elegant linework, and calm spatial rhythm. Weather treatment: Express the sunny condition through warm lake light, translucent gold haze, glimmering water, and a clear sun disk integrated with the layout. The weather should feel grown from Hangzhou's scenery, not pasted on top. Composition: Center a large elegant temperature number, place the city name HANGZHOU as the main title, add date, condition, high/low, humidity, UV, and wind as small premium data chips. Use a graceful vertical composition with a lake horizon, botanical foreground, and generous cream margins. Material and texture: Fine paper grain, silk-like mist, delicate ink outlines, soft enamel weather icons, subtle gilded edges. Color: lake green, ivory, tea-gold, pale blue, warm sunlight amber. Typography and labels: Use short readable English labels only: HANGZHOU, SUNNY, 27°C, HIGH 30, LOW 20, HUMIDITY 48%, UV 7, SE WIND. Keep text clean, premium, and not crowded. Avoid: generic app UI, flat emoji icons, random landmark collage, crowded forecast tables, tiny unreadable numbers, cheap tourism poster style, real logos, watermarks, distorted text, low-resolution artifacts, and visual noise. Set aspect ratio to 4:5.

Open template
Creative Templates: Mechanical Keyboard Exploded View. GPT Image 2 scene prompt preview image.
1 prompts

Mechanical Keyboard Exploded View

Create a 4:5 exploded assembly technical drawing for a compact 75 percent mechanical keyboard with aluminum top case, gasket mount, FR4 plate, hot-swap PCB, switches, stabilizers, keycaps, foam layers, and screws. Scene fit: This scene should be precise, technical, and elegant. Complexity is welcome only when it improves part readability and assembly logic. Visual direction: keyboard parts including shell, plate, switches, keycaps, PCB, and stabilizers. Subject treatment: Break a compact 75 percent mechanical keyboard with aluminum top case, gasket mount, FR4 plate, hot-swap PCB, switches, stabilizers, keycaps, foam layers, and screws into believable components, preserving real assembly order, scale, material differences, and connection logic. Every floating part should feel mechanically necessary and aligned to the central assembly axis. Composition: vertical isometric stack, final silhouette below, callout rails. Use an isometric or clean orthographic exploded view, aligned axes, even spacing, numbered callouts, and one ghosted assembled silhouette for context. Color and material: off-white drafting paper, graphite, muted blue, plastic and metal cues. Use refined lighting, controlled contrast, and coherent material texture that matches this scene. Typography and labels: Use short readable part labels and numbers only. No dense manual paragraphs or fake specifications. Quality rules: The subject and background must form one intentional visual system. Make the prompt result match the template scene first: use ornate detail only where the scene calls for it, and use simplicity where clarity matters. Avoid random floating parts, impossible assembly order, fake logos, messy repair bench, too many screws, illegible microtext, real logos, watermarks, distorted anatomy, low-resolution artifacts, cheap stock clichés, and visual noise. Set aspect ratio to 4:5.

Open template
Infographics: City Food Walk Illustrated Map. GPT Image 2 scene prompt preview image.
1 prompts

City Food Walk Illustrated Map

Create a 9:16 illustrated travel map poster for Hangzhou West Lake food walk from Hubin Road to Hefang Street and Longjing Village, with stops for lotus root starch, scallion noodles, West Lake vinegar fish, tea desserts, and osmanthus cake. Scene fit: This scene should be decorative, useful, and collectible. It must be a clear illustrated map, not a GPS screenshot or messy postcard collage. Visual direction: food-walk route with local dishes and stops. Subject treatment: Represent this specific route as a collectible illustrated map: Hangzhou West Lake food walk from Hubin Road to Hefang Street and Longjing Village, with stops for lotus root starch, scallion noodles, West Lake vinegar fish, tea desserts, and osmanthus cake. Landmarks, streets, food, nature, and culture stops should be simplified into coherent local icons that feel drawn for this place rather than generic travel clip art. Composition: simplified route line, 6-8 food stops, large city title. Use a vertical map, highlighted route, 5-8 stops, large place title, small legend, and readable spacing between icons. Color and material: cream paper, ink outlines, tomato red, jade green, golden accents. Use refined lighting, controlled contrast, and coherent material texture that matches this scene. Typography and labels: Use short readable place labels, stop numbers, and legend words only. No long itinerary paragraphs. Quality rules: The subject and background must form one intentional visual system. Make the prompt result match the template scene first: use ornate detail only where the scene calls for it, and use simplicity where clarity matters. Avoid real map tiles, fake logos, too many street names, unreadable tiny labels, generic tourism icons, cluttered roads, real logos, watermarks, distorted anatomy, low-resolution artifacts, cheap stock clichés, and visual noise. Set aspect ratio to 9:16.

Open template
Fashion & Beauty: Spring Korean Streetwear Guide. GPT Image 2 scene prompt preview image.
1 prompts

Spring Korean Streetwear Guide

Create a 4:5 Korean streetwear styling guide for a fictional young adult model with shoulder-length black hair, neutral body proportions, and a relaxed Seoul street pose, styled for a fresh spring day. Scene fit: This scene should be stylish, clean, and practical for outfit planning. It must not become a chaotic closet collage or fake shopping ad. Visual direction: light spring layering with fresh color and soft casual structure. Subject treatment: Design a complete outfit guide for a fictional young adult model with shoulder-length black hair, neutral body proportions, and a relaxed Seoul street pose, styled for a fresh spring day. Use light trench jacket, white tee, straight denim, slim sneakers, mini shoulder bag, and butter-yellow knit accent. Keep the figure fictional, stylish, and non-celebrity; show one clear full-body look with practical item callouts and a cohesive Korean streetwear mood. Composition: main figure, three item callouts, color swatches, spring cue. Use one main full-body look, 3-4 item callouts, palette chips, and one short styling note with generous spacing. Color and material: cream, denim blue, soft gray, butter yellow. Use refined lighting, controlled contrast, and coherent material texture that matches this scene. Typography and labels: Use short readable item labels only: jacket, tee, denim, sneakers, bag, knit, coat. No prices, fake brands, or shopping links. Quality rules: The subject and background must form one intentional visual system. Make the prompt result match the template scene first: use ornate detail only where the scene calls for it, and use simplicity where clarity matters. Avoid fake brand logos, tiny shopping text, too many outfits, crowded wardrobe boards, identity drift, runway chaos, real logos, watermarks, distorted anatomy, low-resolution artifacts, cheap stock clichés, and visual noise. Set aspect ratio to 4:5.

Open template
Creative Templates: Cozy Clay Bedroom Miniature. GPT Image 2 scene prompt preview image.
1 prompts

Cozy Clay Bedroom Miniature

Create a 4:5 soft clay miniature diorama of a cozy rainy bedroom with a rounded bed, tiny cat lamp, half-open book, slippers, window raindrops, and a warm bedside glow. Scene fit: This scene should be tactile, cute, simple, and handmade. It should not look like realistic product photography or a dense model set. Visual direction: cozy bedroom or home-life clay charm. Subject treatment: Build the scene around a cozy rainy bedroom with a rounded bed, tiny cat lamp, half-open book, slippers, window raindrops, and a warm bedside glow. Every object should look sculpted from soft clay with rounded edges, slight fingerprints, matte surfaces, and toy-like proportions. Keep the scene small, charming, and readable at a glance. Composition: small room box, one character/object, three props, warm shadow. Use one small diorama base, one hero object or character, 3-5 supporting props maximum, clean front or isometric view, and soft studio lighting. Color and material: pastel clay, cream base, soft pink, muted blue. Use refined lighting, controlled contrast, and coherent material texture that matches this scene. Typography and labels: No readable text. Tiny decorative shapes are allowed if they do not become labels or logos. Quality rules: The subject and background must form one intentional visual system. Make the prompt result match the template scene first: use ornate detail only where the scene calls for it, and use simplicity where clarity matters. Avoid photoreal furniture, glossy plastic, crowded dollhouse clutter, hard realistic textures, brand packaging, tiny signage, real logos, watermarks, distorted anatomy, low-resolution artifacts, cheap stock clichés, and visual noise. Set aspect ratio to 4:5.

Open template
Posters & Typography: Three Polaroid City Memory Poster. GPT Image 2 scene prompt preview image.
1 prompts

Three Polaroid City Memory Poster

Create a 9:16 travel memory poster for Hangzhou, using April 12-14, 2026, three photo moments: West Lake boat at dusk, Longjing tea village, Hefang Street snack stall; route through Hubin Road, Su Causeway, Longjing Village, and Hefang Street. Scene fit: This scene should feel personal, elegant, and emotionally specific. It should not look like a generic tourism ad or messy scrapbook dump. Visual direction: classic three-photo city memory with calm rhythm. Subject treatment: Use Hangzhou, April 12-14, 2026, three photo moments: West Lake boat at dusk, Longjing tea village, Hefang Street snack stall; route through Hubin Road, Su Causeway, Longjing Village, and Hefang Street, ticket fragments, route marks, food cues, and three image-like memory modules as the emotional structure of one finished trip poster. The three photo moments should be illustrated or photo-real modules inside the poster, so no external upload is required for this sample image. Composition: three polaroids, city title, date, small route line, stamp details. Use three polaroid or photo modules, a city title, date, small route line, ticket or stamp details, and generous negative space. Color and material: warm cream paper, muted city accents, soft black ink. Use refined lighting, controlled contrast, and coherent material texture that matches this scene. Typography and labels: Use city name, date, and 2-4 short place labels only. No long diary paragraphs or fake travel agency copy. Quality rules: The subject and background must form one intentional visual system. Make the prompt result match the template scene first: use ornate detail only where the scene calls for it, and use simplicity where clarity matters. Avoid messy sticker pile, too many photos, fake airline logos, generic landmark collage, unreadable itinerary text, cheesy quote overlays, real logos, watermarks, distorted anatomy, low-resolution artifacts, cheap stock clichés, and visual noise. Set aspect ratio to 9:16.

Open template
Page 1 / 9
Back to GPT Image 2