GPT Image 2 Prompts
浏览一个持续增长的 GPT Image 2 Prompts 提示词库,获取可复制提示词、来源链接和示例图片,涵盖人像、海报、角色、UI 截图和对比案例。
GPT Image 2 提示词库
浏览带有来源链接和示例图片的可复制 GPT Image 2 Prompts。卡片流负责持续承载新增案例,正文部分则帮助搜索引擎和用户更快理解这些提示词的用途。
Luxury Real Estate App UI Mockup
{"type":"mobile real estate app showcase","style":"high-fidelity modern fintech-style UI mockup, premium luxury aesthetic, deep navy blue gradient backgrounds, glowing cobalt highlights, glassmorphism cards, subtle inner shadows, rounded corners, clean white sans-serif text, realistic iPhone-style device frames shown front-on","brand":{"name":"{argument name=\"brand name\" default=\"PrimeHaus\"}","subtitle":"IMÓVEIS"},"layout":{"canvas":"three smartphone screens side by side on a light neutral background","count":3,"sections":[{"title":"home screen","position":"left","count":7,"labels":["top brand header with notification icon","hero headline","search bar","4 action shortcuts","featured properties heading with see-all link","featured property card","bottom tab bar"]},{"title":"search results screen","position":"center","count":6,"labels":["top back button and search field","3 filter chips plus filter button","results count text","4 vertical property cards","favorite icons on cards","bottom tab bar"]},{"title":"property details screen","position":"right","count":6,"labels":["top back share and favorite icons","large hero image","property info card","feature row","description block","2 call-to-action buttons plus bottom tab bar"]}]},"screens":{"left":{"time":"9:41","headline":"{argument name=\"headline text\" default=\"Encontre o imóvel dos seus sonhos\"}","search_placeholder":"Buscar por localização, bairro ou código","shortcut_buttons":{"count":4,"labels":["Comprar","Alugar","Lançamentos","Favoritos"]},"section_title":"Imóveis em destaque","link_text":"Ver todos","featured_card":{"badge":"Destaque","title":"Casa em Alphaville","location":"Santana de Parnaíba, SP","features":{"count":3,"items":["4","5","320 m²"]},"price":"R$ 3.450.000"},"bottom_tabs":{"count":5,"labels":["Início","Buscar","Favoritos","Mensagens","Perfil"]}},"center":{"search_placeholder":"Buscar imóveis","filter_chips":{"count":3,"labels":["Comprar","Tipo de imóvel","Preço"]},"results_text":"128 imóveis encontrados","cards":{"count":4,"items":[{"title":"Apartamento","location":"Vila Nova Conceição, São Paulo - SP","features":["3","4","178 m²"],"price":"R$ 2.980.000"},{"title":"Casa em Condomínio","location":"Alphaville, Santana de Parnaíba - SP","features":["4","6","450 m²"],"price":"R$ 5.600.000"},{"title":"Apartamento","location":"Itaim Bibi, São Paulo - SP","features":["2","3","112 m²"],"price":"R$ 1.850.000"},{"title":"Cobertura Duplex","location":"Moema, São Paulo - SP","features":["4","5","280 m²"],"price":"R$ 4.750.000"}]},"bottom_tabs":{"count":5,"labels":["Início","Buscar","Favoritos","Mensagens","Perfil"]}},"right":{"image_counter":"1/24","title":"{argument name=\"property title\" default=\"Casa em Alphaville\"}","location":"Santana de Parnaíba, SP","price":"{argument name=\"price\" default=\"R$ 3.450.000\"}","features":{"count":4,"items":["4 suítes","5 banheiros","320 m²","2 vagas"]},"description_title":"Descrição","description":"{argument name=\"description text\" default=\"Casa de alto padrão com projeto arquitetônico moderno, ambientes integrados, iluminação natural e acabamentos de primeira linha. Localizada em um dos condomínios mais valorizados de Alphaville.\"}","more_link":"Ver mais","buttons":{"count":2,"labels":["Contato","Agendar visita"]},"bottom_tabs":{"count":5,"labels":["Início","Buscar","Favoritos","Mensagens","Perfil"]}}},"imagery":{"property_photo_style":"luxury contemporary mansion at dusk with warm interior lighting, expansive glass walls, angular architecture, landscaped yard, large reflective swimming pool, photorealistic real estate photography","card_thumbnails":{"count":5,"style":"high-end residential interiors and exteriors with consistent blue-toned color grading"}},"color_palette":{"primary":"deep navy","secondary":"royal blue","accent":"electric blue","text":"white and light gray"}}
3D Papercut Pirate Scene
平视视角,正面直视一个高饱和度3D多层纸雕场景。 整体为灰蓝、奶油白、深森林绿,点缀琥珀色暖光。所有元素必须是实体纸张分层结构,边缘为硬切割效果,并清晰可见厚纸横截面。 极致丰富构图,一个高能量的{argument name="场景主题" default="海盗群像场景"},发生在混乱的海上冒险中。多个风格化海盗角色以夸张动作呈现,中层核心是一位{argument name="核心角色" default="戴草帽的年轻船长"}向前冲刺,周围是姿态鲜明的船员。 分层的船只、翻涌海浪和旋转洋流贯穿前中后景,多艘船 in 不同空间层中碰撞倾斜,船帆破损、旗帜飞扬。 背景层中出现体型巨大的对立角色剪影,增强压迫感和空间层级。 前景堆满叙事元素:宝箱、金币、地图、木桶、绳索、灯笼,以及一张带笑脸海盗的通缉令。角落放置一个吉祥物式小角色作为视觉锚点。 所有氛围元素以纸雕形式呈现:海浪、水花、烟雾、碎片、动势线,强化混乱与动感。 主体位于中层,元素跨层互动,通过重叠和空间关系表达运动感。 单一方向柔光,形成清晰层间投影。整体为哑光纸张质感,无发光、无柔边、无体积光。 octane渲染,8K,细节清晰 --ar 3:4 --stylize 450
VTuber Character Reference Sheet
{"type":"anime VTuber character reference sheet","style":"clean polished Japanese character design document, white background, pale blue ornamental border, elegant fantasy-idol presentation, high-detail anime rendering, soft cel shading, crisp line art, delicate sparkles and star motifs","subject":{"name":"{argument name=\"character name\" default=\"蒼空なつみ\"}","romanized_name":"SORA NATSUMI","age":"18","school_year":"high school senior","birthday":"July 7","height":"145 cm","theme":"sky, stars, clouds, sailor-inspired magical idol","hair":{"color":"{argument name=\"hair color\" default=\"pastel sky blue gradient\"}","style":"very long wavy twin-tail-like side sections with a rounded crown loop bun, soft bangs, gradient ends, star accessories, long back hair flowing to the waist"},"eyes":{"color":"{argument name=\"eye color\" default=\"blue\"}","style":"large gentle anime eyes with bright highlights"},"outfit":{"silhouette":"short layered idol dress with sheer draped overskirt panels","pieces":["small sailor hat tilted to one side","white off-shoulder blouse with voluminous sleeves","large navy chest ribbon with gold star brooch","navy corset-style bodice with gold trim","short pleated layered skirt in navy and blue","translucent light-blue side and back drapes with constellation pattern","star charm waist compass ornament","blue choker with star charm","thigh garter on one leg","frilled ankle socks","white lace-up ankle boots with blue gradient and gold details"]}},"layout":{"canvas":"landscape reference sheet","sections":[{"title":"FRONT","position":"upper center-left","count":1,"labels":["full-body front view"]},{"title":"SIDE","position":"upper center","count":1,"labels":["full-body side view"]},{"title":"BACK","position":"upper center-right","count":1,"labels":["full-body back view"]},{"title":"ディテール&パーツ","position":"upper right","count":9,"labels":["胸元ブローチ(拡大)","ウエストコンパス(拡大)","袖の刺繍(拡大)","ヘアクリップ(拡大)","帽子(サイド)","背面リボン(拡大)","まももベルト(拡大)","靴デザイン(拡大)","靴底(拡大)"]},{"title":"表情バリエーション","position":"lower left-middle","count":6,"labels":["通常","笑顔","照れ","怒り","悲しみ","ウィンク"]},{"title":"カラーイメージ","position":"bottom left","count":5,"labels":["#7EC6FF","#B7E3FF","#1C2A6B","#FFFFFF","#FFD97A"]},{"title":"モチーフ・イメージ","position":"bottom center-left","count":5,"labels":["空","雲","星","星座","コンパス"]},{"title":"設定小物・備品","position":"mid-right lower","count":5,"labels":["星雲モチーフランプ","お供のぬいぐるみ『くもりん』","星空タブレット","魔法のステッキ","ドリンクボトル"]},{"title":"ロゴデザイン(活動名義)","position":"bottom center","count":1,"labels":["蒼空なつみ / SORA NATSUMI"]},{"title":"Live2D・その他設定","position":"bottom right","count":1,"labels":["notes block"]}],"left_info_panel":{"count":10,"labels":["年齢","誕生日","身長","体重","血液型","所属","活動内容","ファンネーム","配色タグ","切り抜きタグ"]}},"composition":"three large full-body views dominate the center, surrounded by neatly boxed information panels and detail callouts with fine Japanese annotation lines","text_treatment":"most labels and notes in elegant Japanese editorial typography with blue headings and serif accents","color_palette":"{argument name=\"primary color palette\" default=\"sky blue, navy blue, white, and pale gold\"}","quality":"high resolution, professional character sheet, visually balanced, extremely detailed costume design, suitable for VTuber branding and Live2D model planning"}
Celestial VTuber Livestream Overlay
{"type":"anime VTuber livestream overlay illustration","theme":"celestial night sky, stars, constellations, deep sapphire and white, elegant magical idol aesthetic","subject":{"character":{"name":"{argument name=\"character name\" default=\"蒼空なつみ\"}","gender_presentation":"cute young anime woman","hair":{"color":"{argument name=\"hair color\" default=\"pale sky blue\"}","style":"very long wavy twin-tail-like loose hair with a small ahoge, soft bangs, star hair ornaments and ribbons"},"outfit":{"style":"ornate celestial idol dress","pieces":["white off-shoulder blouse with puffy sleeves","large navy blue chest bow with gold star ornament","navy corset-style bodice with gold trim and constellation motifs","star choker necklace","small tilted white mini hat with blue ribbon and gold star accessory"]},"pose":"seated at a streaming desk, centered, facing forward behind a microphone and keyboard","expression":"gentle, calm, welcoming"}},"environment":{"setting":"cozy fantasy streaming room at night with a large window showing a star-filled sky","background_elements":["hanging star garlands and crystal ornaments","shelves with bottles, books, hourglass, telescope-like objects, and celestial decor","moon and constellation imagery outside the window","two fluffy mascot-like cloud plushies with blue eyes"]},"props":{"desk_items":["1 large vintage-style microphone on a stand at left of center","1 decorative keyboard with star-themed keys","1 nameplate reading \"蒼空なつみ - SORA NATSUMI -\"","1 parfait glass dessert topped with a small white bunny figure","1 small cup or candle holder with a cloud mascot"]},"layout":{"format":"full livestream waiting/talking screen overlay","sections":[{"title":"LIVE","position":"top-left vertical banner","count":3,"labels":["LIVE","20:30","05/20 Tue"]},{"title":"COMMENTS","position":"left panel","count":14,"labels":["星見たこ:こんなっつー! ✨","夜空ヨル:待機完了〜!今日はどんなお話かな?","しろくまアイス:こんなっつー!","天の川りゅうせい:今日も可愛い!大好き!","ふるーすたー:も ふ ぷ","月見うさぎ:楽しみ〜!!","ななせそら:こんなっつー! ☁️✨","Stella:きちゃー!( *´ω`)","彗星カケル:配信ありがとう〜!","みずいろねこ:こんなっつー! 💙","星屑ひかり:今日のサムネもかわいい…!","夜更かしの民:こんなっつー!","ゆめみがち:待ってました!","ハル:今日の衣装も最高に似合ってるぞ〜!"]},{"title":"social tags","position":"below comments panel","count":2,"labels":["@sora_natsumi","#なつみんらいぶ"]},{"title":"goal panel","position":"bottom-left","count":4,"labels":["目標:チャンネル登録者数 100,000人!","現在","58,320 人","いつもありがとう!チャンネル登録高評価よろしくね☆"]},{"title":"supporters panel","position":"bottom-left small card beside goal panel","count":6,"labels":["☆今日の星屑さん(スパチャありがとう!)☆","星の旅人 さん ¥2,000","ふるーすたー さん ¥1,000","天の川りゅうせい さん ¥500","みずいろねこ さん ¥320","流れ星 さん ¥200"]},{"title":"Now Talking","position":"upper-right panel","count":3,"labels":["Now Talking","{argument name=\"talk topic text\" default=\"初見さん歓迎!雑談配信✨\"}","{argument name=\"subtitle text\" default=\"〜まったりお話ししよっ〜\"}"]},{"title":"SET LIST","position":"right-middle panel","count":5,"labels":["01. 星屑になれたら","02. プラネタリウム","03. 夜に駆ける","04. Stellar Stellar","05. 天体観測"]},{"title":"promo card","position":"bottom-right panel","count":3,"labels":["オリジナル曲","星空コンパス","MV公開中!"]}],"logo":{"position":"top-right corner","text":"蒼空なつみ SORA NATSUMI"}},"style":{"rendering":"high-detail polished anime illustration, luminous highlights, soft glow, crisp linework","composition":"symmetrical, dense decorative overlay, character centered with UI panels around her","lighting":"cool moonlit blue ambiance with sparkling gold accents","mood":"dreamy, magical, cozy, elegant"}}
AI Chat Screenshot with Anatomy Poster
A smartphone screenshot of an AI chat app interface in portrait orientation, showing a clean white iOS-style screen with a top bar containing a hamburger menu on the left, the blue title “Thinking” with a small chevron, and two icons on the right for compose and more options. Near the top is a rounded light-gray user message bubble in Japanese reading “Attention機構を人体に喩えて細かくラベルを付けて。医学書の精緻な挿絵風に。” Below it, faint gray system text reads “Thought for 5s” and “画像が作成されました”. The lower half of the screen displays the generated image as a detailed medical textbook-style anatomical poster. The poster shows a realistic three-quarter back view of an adult woman from head to upper thighs, with her head turned slightly left and her dark brown hair tied into a messy bun. Her skin and outer tissues are removed across most of the body to reveal musculature, nerves, blood vessels, and fibrous structures, rendered in extremely detailed vintage medical illustration style on a warm cream paper background. A rectangular cutout obscures the center of the face from forehead to jaw, creating a censored blank block over the facial area. Fine glowing network lines and nodes are overlaid through the exposed brain and body to visualize the attention mechanism as if it were an anatomical system. Surrounding the figure are 6 numbered explanatory panels and 2 additional text blocks, all densely filled with Japanese and English bilingual labels, formulas, arrows, and diagrams. The main poster title at upper left is “アテンション機構の解剖学的構造” with the English subtitle “The Anatomy of the Attention Mechanism”. At the top right is a flowchart labeled “Overall Flow of Attention” containing 5 connected boxes describing the process from input to output. Down the left side are 3 stacked explanatory sections with equations and mapping tables. Down the right side are 3 more numbered sections, each with technical notes and a small chart or matrix; the visible mini-diagrams include 1 purple heatmap matrix, 1 blue bar chart, and 1 multicolored cube graphic. Along the bottom edge are 5 small brain inset illustrations with captions showing stages of information flow. The overall aesthetic should feel like a highly polished AI-generated screenshot documenting a surreal scientific illustration request, with crisp UI, lots of tiny readable labels, and a convincing fusion of transformer attention diagrams and classical anatomical plate design.
AI Chat Screenshot with Cosmic Artwork
A vertical smartphone screenshot of an AI chat app interface in light mode, showing the moment an image has just been generated. At the very top is a simple mobile app header with a hamburger menu icon on the left, the blue title “Thinking” with a right chevron beside it, and on the right a compose icon and a three-dot menu. Below that is the bottom portion of a rounded light-gray user message bubble containing Japanese text, visible across 4 lines: 「水彩や線画、油彩、カラー版画、 金箔アクセントなどを取り入れ てテクスチャも入れて立体的に、 芸術爆発させて」. Under the bubble, on a white background, place the gray system status line “Thought for a couple of seconds”, then a darker gray Japanese line “画像が作成されました”. Below that, centered and taking most of the lower half of the screen, display a large rounded-corner generated artwork preview. The artwork itself is an ultra-detailed cosmic visionary collage in a square format: a tiny glowing human figure stands at the vanishing point in the center bottom-middle, radiating light; two large semi-transparent human profiles or heads face inward from left and right midground; swirling galaxies, stars, nebulae, planets, floating geometric frames, transparent spheres, gold-leaf fragments, line art, splattered paint, and layered textures fill the scene. Use a dramatic palette of deep indigo, ultramarine, violet, magenta, white, and luminous gold, with explosive painterly energy combining watercolor, oil painting, printmaking, ink lines, and metallic foil accents. Include branching tree-like forms, abstract architectural steps on the right, a lotus-like flower on the left, multiple framed mini-scenes embedded in space, and dense luminous particle trails converging toward the central figure. The overall screenshot should feel like a real iPhone app capture, clean and minimal outside the artwork, with the generated image being the visual focal point.
Mobile AI Chat Screenshot with Transformer Poster
A smartphone app screenshot in portrait orientation showing an AI chat interface in Japanese, with a clean white background and iOS-style top bar. At the very top, display a left hamburger menu icon, blue text reading “Thinking” with a right chevron, an edit compose icon, and a three-dot menu. Below it, show a large rounded light-gray chat bubble containing 6 lines of Japanese text, exactly: 「モーフ感浮遊感のある感じ+かつ反射するダストのレイヤーを加えて。テキスト多め、重要ーアーキテクチャは可読性高め、精細で細かいコードや数式の可読性は二の次のデザインアクセント&アートエレメンツで。」 Under the bubble, place small light-gray system text reading “Thought for a couple of seconds”, then a larger gray Japanese line reading “画像が作成されました”. In the lower half, show 1 generated image thumbnail with rounded corners: a dense futuristic infographic poster about the Transformer neural network architecture, rendered as glowing neon sci-fi holographic UI art. The poster should have a cosmic deep-space background with stars, purple, magenta, blue, and cyan light trails, reflective particles, volumetric glow, and floating geometric crystals. Large title at top left of the poster: “TRANSFORMER”, with subtitle “THE ARCHITECTURE”. Include 2 large central glowing boxes labeled “ENCODER LAYER” and “DECODER LAYER”, linked by arrows and smaller blocks for “ENCODER STACK (N×)” and “DECODER STACK (N×)”. Add many surrounding labeled panels and formulas, including sections for “KEY INNOVATIONS”, “WHY TRANSFORMERS?”, “POSITIONAL ENCODING”, “FEED FORWARD NETWORK”, “COMPUTATIONAL COMPLEXITY”, “OVERALL FLOW”, “SELF-ATTENTION MECHANISM”, “MULTI-HEAD ATTENTION”, “APPLICATIONS”, and “MODEL SCALING”. Show token rows at the bottom, including an English input sequence like “[CLS] the cat sat on the mat [SEP]” and an output sequence in French-like tokens ending with “[EOS]”. Add tiny charts, equations, attention diagrams, and wiring lines throughout, but keep the overall impression artistic and luminous rather than strictly readable. At the bottom center of the poster thumbnail, overlay a white circular download button with a black downward arrow. The overall screenshot should feel like a real mobile ChatGPT-style interface displaying an AI-generated image result, crisp and modern.
AI Thinking App Screenshot
A smartphone chat app screenshot in portrait orientation, minimal iOS-style interface on a white background. At the top is a navigation bar with a hamburger menu icon on the left, the blue title “Thinking” near the top center-left with a right chevron, and 2 icons on the right: a compose/edit square icon and a horizontal three-dots menu icon. Below it is a large rounded light-gray chat bubble containing 2 lines of Japanese text: “AIが見ている世界を抽象的かつ” and “幻想的に画像で表現して”. Under the bubble, left-aligned gray system text reads “Thought for 7s”, and below that another gray Japanese line reads “画像が作成されました”. The lower half of the screen shows a single generated image card with large rounded corners. The image is a surreal abstract cosmic data landscape: a glowing white horizon at the center, a dark reflective grid floor receding in deep perspective, and explosive streaks of light, nebula clouds, stars, and particles rushing outward from the vanishing point. Floating throughout the space are dozens of translucent rectangular image panels or data windows, some faint and some bright, suspended at different depths. In the upper right area of the artwork, include a luminous network node diagram made of connected points and lines. Dominant colors are electric blue, indigo, violet, cyan, and touches of magenta and gold. The mood is futuristic, vast, dreamlike, and intelligent, as if visualizing the world as seen by an AI. Clean mobile app screenshot composition, realistic UI, crisp typography, high-contrast luminous digital art in the embedded image.
Dark Manga Monster Poster
A dark monochrome manga-style horror illustration on a pure black background, in a tall vertical poster composition. Centered at the bottom is a small lonely teenage boy with messy short dark hair, sitting on the ground with his knees pulled to his chest and his arms wrapped around them, head slightly bowed, conveying fear, isolation, and helplessness. Towering over him is a gigantic biomechanical demon or armored monster head descending from above, with a long angular skull-like face, sharp chin, narrow hollow eye sockets, layered helmet-like plates, and a jagged open mouth. Surrounding the boy are exactly 2 enormous skeletal clawed hands, one on the left and one on the right, framing him like a cage; each hand has long segmented metallic fingers ending in sharp pointed tips, bent inward menacingly. Render everything in rough scratchy white ink linework and heavy crosshatching against deep black, with intense high-contrast seinen manga energy. Add vivid neon purple paint-like liquid in splashes and long dripping streaks from the monster’s face, the hands, and the ground, pooling around the boy for the only strong color accent. On the upper right side, add vertical Japanese text reading 逃げちゃダメだ, and beneath it in small white English text: EGO BORDER: ACTIVE. The mood is oppressive, psychological, claustrophobic, and nightmarish, like a dark anime poster about inner fear and mental collapse.
Gritty Evangelion Unit-01 Poster
A gritty vertical anime-mecha poster in distressed screen-print style, inspired by retro military dossiers and Evangelion key art. The main subject is a massive close-up side profile of {argument name="mecha unit name" default="EVANGELION: UNIT-01"}, dominating the right two-thirds of the composition, facing left with its jaws open in a feral scream. The head is angular and biomechanical, with deep purple armor, olive-green accent plates, black metal inner structures, a single long horn projecting diagonally upward to the left, sharp teeth, and a narrow glowing eye. The surface is scratched, chipped, weathered, and inked with heavy comic-style linework. Behind it is a dark navy blueprint background filled with 8 visible technical diagram clusters, white schematic lines, anatomical overlays, cross-sections, labels, measurement marks, and interface graphics. A violent red paint slash explodes diagonally across the upper center and left side, with additional red drips running downward from the mecha’s mouth area. In the lower left corner, 1 small human figure sits hunched on a gray equipment crate: a teenage boy in a short-sleeve white school shirt, dark pants, and white sneakers, elbows on knees, looking down at a glowing handheld device, conveying isolation and anxiety. Add layered poster typography and warning graphics throughout the border: a torn yellow caution sign in the upper right with Japanese warning text and “CAUTION NERV ONLY”; a tall left sidebar with “01”, barcode-like blocks, hazard icons, and red-black labels; large distressed white kanji vertically stacked over the mecha torso; the word {argument name="main title" default="ESCAPE"} in spaced red capitals beneath the kanji; a small black label reading “PILOT: {argument name="pilot name" default="IKARI, SHINJI"}”; a bottom-right NERV logo; and footer text such as classified file markings and top secret stamps. Include 6 clearly readable English technical labels in the blueprint area: “TEST TYPE”, “BIOLOGICAL CONSTRAINT”, “AT FIELD”, “MAGI SYSTEM”, “INTERNAL”, and “CLASSIFIED MATERIAL”. Add Japanese dossier text blocks as authentic graphic texture. The palette is limited to dark indigo, black, crimson red, purple, olive green, dirty beige, and off-white. High contrast, torn-paper collage details, halftone texture, ink splatter, cracked paint, and worn print edges. Cinematic, oppressive, chaotic, collectible poster design, ultra-detailed, sharp, textured, no clean modern polish.
Cosmic Ink Dragon Roar
A dramatic vertical close-up of a roaring cosmic dragon head in side profile, painted in a violent ink-splatter and brushstroke style on a stark white background. The creature has a long sharp horn, angular armored facial plates, a glowing white-green eye, and a huge open mouth lined with jagged teeth. Inside the mouth and throat, instead of flesh, there is a swirling galaxy-like void of luminous {argument name="energy color" default="violet"} and magenta nebula light, with sparkling stars and dust erupting outward like cosmic fire. The dragon’s body is mostly black and deep indigo with glossy highlights, accented by iridescent emerald reflections along the neck and scales. Use expressive high-contrast sumi-e-inspired strokes, splattered paint textures, torn abstract edges, and explosive particle effects concentrated around the eye, mouth, and upper head. Composition should fill the frame with the head tilted upward in a feral scream, cropped tightly for maximum intensity, with no environment and no extra subjects, just raw abstract energy, ink, and celestial debris.
GPT Image 2 Prompts 常见问题
这些 GPT Image 2 Prompts 应该怎么使用?
先复制一个最接近目标风格的提示词,再替换主体、色彩、镜头语言或构图细节。相比每次从零开始,基于一个强提示词做定向改写,通常更容易得到稳定结果。