Image Prompts for Comic Graphic Novel

このキャラクターと画風を維持して、プロのデザイナーが制作したような、いろいろな{argument name="テーマ" default="ライブ衣装やポーズのライブ告知チラシ"}を2×2の4分割で出して。 衣装、ポーズ、音楽ジャンル、背景、照明、チラシのデザインは自由に発想し、4つそれぞれに違った魅力と意外性を持たせて。ライブの熱気や世界観が伝わるビジュアルと印象的なタイポグラフィを取り入れ、公演タイトルや日時・会場などの架空の告知情報も含めて、プロ級の仕上がりにして。 4つすべて同じキャラクター、同じ画風を維持する。全体の比率は4:3で。

“{argument name="セリフ" default="ケンカなんかしないよ？"}”という男の子と、その吹き出しに噛みつく凶暴そうな女の子。アスペクト比は縦長（4:5）で描画

Goal: Create a wide side-by-side image showing a rough manga storyboard on the left and a fully rendered black-and-white manga page on the right, demonstrating the transformation from crude layout sketch to polished fantasy comic art. Canvas: Horizontal 3:2 composition, white background, two pages placed next to each other with a small gap. Left page occupies about 45% width, right page about 55% width. Left page: A crude storyboard drawn like a marker sketch on a white sheet with thick black panel borders. Use bright magenta and purple stick-figure drawings and rough silhouettes. Include exactly 6 rough panels: 1) top-left panel with a magenta stick figure standing, 2) top-center narrow panel with a large black oval speech bubble, 3) top-right vertical panel with purple monster-like shapes and sound-effect scribbles, 4) middle-wide panel with a magenta figure lying on the ground, 5) bottom-left panel with a purple hooded figure silhouette and an oval speech bubble, 6) bottom-right panel with a magenta kneeling or fallen figure and a large black oval speech bubble. Keep the draft intentionally messy, childlike, and simplified, with visible Japanese manga sound effects and dialogue placement. Right page: A polished monochrome manga page based on the same layout, high-detail fantasy shoujo style, dramatic ink work, screentones, speed lines, cinematic panel angles, heavy black gutters, snowy or ruined gothic background. Include exactly 7 finished manga panels: 1) top-left wide panel showing a small blonde girl in an ornate fluffy rabbit-ear hood turning in fear inside a ruined stone city, with a vertical speech bubble, 2) top-right panel showing a menacing hooded monster with a toothy dinosaur-like hood looming behind her as she screams, 3) middle-left wide panel showing the rabbit-hooded girl collapsed face-down on the ground with long floppy ears spread out, 4) middle-right small panel showing a sharp clawed hand or blade dripping dark liquid, 5) bottom-left tall panel showing a calm sinister girl wearing a monster-mouth hood and ornate dark fantasy cloak, looking down with a smug expression, 6) bottom-center/right tall panel showing the villain standing full-body in a lavish frilled outfit among gothic ruins, 7) bottom-right panel showing the defeated rabbit-hooded girl lying motionless on wet stone. Characters: The victim is a cute young anime girl with wavy light hair, big eyes, a fluffy bunny hood with long ears, frilled dress, ribbons, lace, and soft innocent expression. The antagonist is a similar young anime girl but darker and more threatening, wearing a monster or dragon-mouth hood with sharp teeth, spiked accessories, ornate cloak, frills, and a cold confident smile. Text content: Use Japanese manga typography, vertical speech bubbles, and bold hand-drawn sound effects. Include the key dialogue as Japanese text, such as {argument name="victim dialogue" default="……なんで……お前が……？"}, {argument name="body dialogue" default="身体が……動かない……"}, {argument name="villain reveal dialogue" default="まだ分かんないの？ 最初から、こうするつもりだった"}, and {argument name="final dialogue" default="もう、お前に用はない。"}. Add large sound effects like 「ザッ」「ザクッ」「ドサッ」「スッ」 around action panels. Visual style: Right page should be professional grayscale manga, highly detailed line art, dramatic lighting, gothic fantasy ruins, textured screentone shading, expressive faces, ornate costume details, cinematic horror-fantasy mood. Left page should remain an intentionally primitive planning sketch with thick black borders, uneven lines, simple colored marker figures, and rough speech bubbles. Constraints: The image must clearly contrast rough storyboard versus finished manga. Keep both pages readable, maintain the same general story beats between left and right, use no color on the finished manga page except grayscale, and do not add extra panels beyond the specified counts.

Try generating {argument name="characters" default="Chiikawa and Usagi"} in {argument name="style" default="Jojo style"}

Using the provided reference image, clean up and regenerate the same composition as a higher-quality version. Remove the gritty/noisy texture, compression artifacts, speckling, and rough grain while preserving the original layout, Japanese text, manga illustration style, colors, lighting, characters, objects, borders, and all readable typography exactly as in the reference. Keep the two-part page structure unchanged: the top Japanese text panel and the bottom tavern illustration. Do not redesign, crop, add elements, translate text, or change the scene; only make it smoother, sharper, cleaner, and more polished like a restored high-resolution print. Apply {argument name="cleanup strength" default="strong grain and artifact removal with preserved ink detail"}.

Goal: Create a square illustrated parody video game cover/poster in the style of a Grand Theft Auto V key art collage, themed around {argument name="series theme" default="Breaking Bad and Better Call Saul in Albuquerque"}. Make it look like polished digital box art with dramatic realism, comic-book painting, bold black panel borders, and a central oversized game logo. Canvas: 1:1 square poster, high resolution, full bleed, saturated desert colors, crisp inked edges, cinematic lighting, no photo collage. Use a black background grid with tilted rectangular panels surrounding the center. Layout: Use exactly 8 illustrated scene panels around one central title logo. The panels are separated by thick black diagonal borders. The central logo overlaps the panels and reads {argument name="main title text" default="grand theft auto"} in large stacked white block letters with black outline, plus a green Roman numeral V emblem with a white ribbon reading {argument name="ribbon location text" default="ALBUQUERQUE"}. Panel count and contents: 1) Top-left large panel: stern bald middle-aged man with glasses, goatee, black porkpie hat, and dark jacket, with a beige RV in the desert behind him; include handwritten text “ALL BAD THINGS MUST PAY”. 2) Top-center panel: young man with stubble wearing a black-and-red beanie and hoodie, looking worried against a pink sunset city background; include handwritten text “YEAH BITCH!”. 3) Top-right narrow panel: helicopter flying over desert mountains and a water tower labeled “ALBUQUERQUE”. 4) Middle-left panel: slick lawyer in a suit pointing finger-guns toward the viewer, yellow legal-ad background; include red script “Better Call Saul!” and small text “ATTORNEY AT LAW”. 5) Middle-right panel: composed man in glasses and a dark suit in front of a fast-food sign reading “LOS POLLOS HERMANOS”; include small tagline “A HIGHER STANDARD”. 6) Bottom-left panel: bald older tough man in a black jacket looking over his shoulder near industrial tracks; include handwritten text “NO HALF MEASURES”. 7) Bottom-center panel: rear view of a black classic car driving toward the Albuquerque skyline and desert mountains, license plate reading “HEISENBRG”, roadside sign reading “ALBUQUERQUE NEW MEXICO”. 8) Bottom-right panel: plastic bag of blue crystal meth beside stacks of cash; include handwritten text “COOK SELL SURVIVE” and a small yellow square Rockstar-style rating/logo icon with a black “R” and white star. Visual style: Highly detailed painterly realism mixed with GTA loading-screen illustration, strong outlines, warm desert yellows, turquoise blues, pink skies, noir shadows, expressive faces, dramatic cropped portraits, clean readable typography. Constraints: Do not use actual actor names. Keep all visible text spelled exactly as specified. Maintain the exact 8-panel collage plus central logo. Avoid extra panels, watermarks, UI elements, or modern app overlays.

Goal: Create a clean monochrome manga expression sheet for a single consistent anime character, showing exactly 9 bust-up portraits in a 3-by-3 grid on a plain white background. Canvas: Vertical portrait canvas, roughly 2:3 aspect ratio, white paper-like background, no borders, no captions, no speech bubbles, no color except grayscale manga shading. Character: A young anime girl named {argument name="character name" default="unnamed expressive manga heroine"}, drawn consistently in every panel. She has {argument name="hair color" default="medium gray"} short wavy bob hair with flared side tufts, a prominent curled ahoge loop rising from the top, a wide black headband, a large pale hair streak on the viewer-left side of her bangs, round cheek marks, large eyes, and small earrings. She wears the same futuristic school-uniform style outfit in all poses: black sleeveless high-collar vest with angular trim, pale long sleeves, shoulder tabs, vertical decorative stripe marks on the chest, a narrow tie-like front panel, and a small circular emblem on the left chest. Layout: Arrange exactly 9 discrete bust portraits, evenly spaced, three columns and three rows. Each portrait is cropped from upper torso to above the hair curl, with slight head angle variation but identical character design and outfit. Keep the drawings separated by white space, like a reference sheet for manga production. Expressions, counted exactly 9: 1. Top left: startled shout, mouth wide open with small fangs visible, eyes wide, sweat drops on cheeks. 2. Top center: anxious grimace, teary half-lidded eyes, uneven clenched mouth, sweat drops. 3. Top right: tired or disappointed side glance, head tilted down, heavy eyelids, small frown. 4. Middle left: skeptical unimpressed look, eyes narrowed to the side, mouth flat, brows low. 5. Middle center: bright cheerful smile, eyes open and sparkling, mouth open happily, relaxed brows. 6. Middle right: surprised alert look, eyes wide and looking upward to the side, small closed mouth, raised brows. 7. Bottom left: frightened gasp, round eyes, open rectangular mouth, sweat drops, worried eyebrows. 8. Bottom center: stern determined glare, eyes narrowed straight forward, eyebrows sharply angled, small serious mouth. 9. Bottom right: crying breakdown, head bowed slightly, eyes squeezed shut, streaming tears, mouth open crying, both hands raised near the face wiping tears. Visual style: Black-and-white Japanese manga line art with soft gray screentone shading, crisp clean outlines, expressive face acting, polished character-sheet presentation, consistent proportions across all nine portraits, no background objects. Constraints: Do not add labels or text. Do not change the character between expressions. Avoid generic emoji-like faces; make each expression fit the same character’s personality. Keep the image as a manga reference sheet rather than a comic panel.

Ultra-detailed blue ballpoint pen illustration translated into monochrome black ink, dense cross-hatching and fine hatch shading, intricate biro sketch texture, obsessive linework, engraved contour rendering, hand-drawn pen illustration aesthetic, expressive organic strokes, editorial surreal composition, high-detail facial rendering, layered sketchbook construction lines, vintage technical scribbles and handwritten notes in pure white only over {argument name="background color" default="bold red background"}, white typographic annotations integrated into backdrop, underground zine poster aesthetic, raw imperfect ink marks, high-contrast black ink subject against vivid red field, experimental graphic design, avant-garde magazine cover style --ar 9:16 --stylize 750.

【プロダクトとコピー】 ・メインコピー: {argument name="メインコピー" default="超軽量・大音量！最新ポータブルスピーカー"} ・擬音・オノマトペ: {argument name="擬音" default="ドォーン!!"} ・吹き出しの中の言葉: {argument name="吹き出し文字" default="サイコー！"} 【画質・演出・構図指定】 ・スタイル: ロイ・リキテンスタインやアメコミのスタイルのような、ドット（網点）のテクスチャと力強い黒いアウトラインが特徴のビンテージポップアートイラスト。 ・被写体: 中央に勢いよく飛び出すように描かれた製品。製品の背後から放射状のライン（集中線）が広がり、躍動感を引き立てている。 ・タイポグラフィと吹き出し（GPT-image / Nano Banana Pro向け）: {argument name="擬音" default="ドォーン!!"}というオノマトペがアメコミ調の立体的でカラフルな文字で大迫力に描かれ、{argument name="吹き出し文字" default="サイコー！"}というテキストが古典的な吹き出し（スピーチバルーン）の中に配置されている。画面上部には{argument name="メインコピー" default="超軽量・大音量！最新ポータブルスピーカー"}のバナーがデザインされている。 ・色彩・トーン: 黄、赤、青、黒といった鮮やかで対比の強いポップアートカラー。 【レイアウトおよび出力上の厳格な制約（必須）】 ・完成したデザインそのものを、画面全体いっぱいに出力してください。デザイン内部の背景やシーン描写（壁・空間・影など）は本文の指示に従って構いません。 ・禁止事項：完成ポスターを額縁に入れた写真、壁に貼った状態の写真、机や紙の上に置いたモックアップ写真、紙の縁の遠近の歪みやドロップシャドウ。 ・Output the finished flat 2D design itself, filling the entire canvas. Scene elements inside the design (walls, rooms, shadows) described above are allowed. Absolutely NO photo-of-a-poster mockups: no picture frames, no poster-on-wall or poster-on-desk shots, no perspective warp or drop shadow around the artwork's edges. ・アスペクト比: 「--ar 1:1」または「--ar 4:3」

Goal: Create a rough Japanese manga storyboard/name page in black and white with hot-pink editor-style sketch marks overlaid, as if planning an action scene in GPT Image 2.5. Canvas: Vertical manga manuscript page, portrait orientation, white background, thin black panel borders, monochrome art with vivid neon pink rough layout lines. Use dramatic Japanese manga pacing with large sound effects and vertical speech balloons. Layout: Use exactly 4 main visible manga areas plus 2 small inset cutaway panels. The page is divided into: 1 large top-left action panel taking about the left half and upper-middle page, 1 top-right reaction/dialogue panel, 1 lower-left small tilted inset panel, 1 lower-right small tilted inset panel, and an open lower central white space containing big dialogue balloons and a pink arrow gesture. Add irregular panel angles and thick black dividing lines. Main panel details: In the large top-left panel, show a crude hot-pink stick-figure runner with an oversized circular head and simple limbs, leaning forward and sprinting. Surround the runner with exactly 6 large distressed black Japanese sound-effect marks reading like 「ド」 and 「ドド」, plus a huge vertical bold shout on the left: 「ぶちかます!!!」 and a smaller vertical line near the center: 「まっすぐ行って!」. The action should feel forceful and unfinished, like a manga name draft with placeholder anatomy. Top-right panel details: Show an almost empty room-like rectangular space with a simple wall line and two hot-pink circular placeholder heads connected by a diagonal line, implying two characters watching. Include exactly 2 vertical speech balloons: one upper balloon reading 「こんなもの……」 and one lower balloon reading 「作戦でも論理でもないことは百も承知だが」. Add exactly 2 small jagged reaction/SFX boxes between the heads, one with a short surprised mark and one with scratchy katakana-style noise. Lower section details: Place exactly 2 large vertical speech balloons in the open lower half. The right balloon reads 「冷静沈着なデータキャラが…」. The center balloon reads 「仲間のためにデータを捨てる展開…!!」. Draw a long hot-pink arrow/body line starting from a large pink circle at lower left and pointing toward the small lower-right inset, with a small black sound effect near the arrow reading 「ゲホッ」. Inset panels: Include exactly 2 grayscale character inset panels. The lower-left tilted inset shows a serious anime woman with long hair and glasses adjusting her glasses, shaded with clean manga hatching, with a vertical caption reading 「バカバカしい根性論など無駄だ」. The lower-right tilted inset shows the same glasses-wearing woman near charts or data displays, speaking in a balloon that reads 「戦いはデータだよ」. The character should look analytical, calm, and data-focused. Visual style: Rough manga storyboard/name, energetic black ink, thick distressed lettering, clean speech balloons, unfinished hot-pink construction lines, minimal backgrounds, high contrast, Japanese weekly manga draft aesthetic. Keep the pink sketch marks visibly on top of the final page art, like an editor's planning layer. Parameters: The analytical character is {argument name="character name" default="the glasses-wearing data strategist"}; the main vertical shout is {argument name="main shout text" default="ぶちかます!!!"}; the action instruction is {argument name="action instruction text" default="まっすぐ行って!"}; the central payoff line is {argument name="payoff dialogue" default="仲間のためにデータを捨てる展開…!!"}; the overlay sketch color is {argument name="sketch color" default="hot pink"}. Constraints: Preserve the exact counts: 4 main visible manga areas, 2 inset panels, 2 grayscale character cutaways, 2 top-right speech balloons, 2 lower speech balloons, 6 large action SFX marks in the top-left panel. Do not add color except the hot-pink sketch layer. Do not add a watermark or page number.

Goal: Create a monochrome Japanese manga page about a quiet office woman gradually opening up after meeting a stray cat, in a sentimental slice-of-life style. Use {argument name="page format" default="vertical manga page, black and white, 4:5 aspect ratio"}. Canvas: Full-page manga layout with clean black panel borders, grayscale screentones, fine ink linework, soft shading, realistic urban backgrounds, and expressive but restrained character acting. The reading direction and text boxes should feel like a Japanese manga page, with vertical Japanese dialogue and narration. Main subject: A young Japanese office woman, {argument name="character name" default="unnamed heroine"}, early 20s, shoulder-length straight black hair with bangs, gentle tired eyes, wearing a cardigan over office clothes and carrying a shoulder bag. She begins lonely and withdrawn, then becomes warmer and more confident. The supporting subject is exactly 1 black-and-white stray cat with round eyes, a fluffy chest, and a slightly scruffy but cute appearance. Layout: Use exactly 9 distinct panels arranged in a dense manga grid. 1. Top-left panel: the woman sits at an office desk, resting her cheek on her hand, looking tired beside a computer monitor and papers. Add a thought bubble and vertical narration. 2. Top-right wide panel: she walks through a city street with buildings and pedestrians behind her, holding her bag strap, looking reserved. Add two tall narration boxes. 3. Middle-left panel: she crouches in an alley or side street and offers food in a small dish to the stray cat. The cat eats while she watches shyly. Add a speech bubble. 4. Middle-right panel: the cat sits alone beside a wall in the alley, looking directly at the viewer. Add a small cat sound effect and a tall narration box. 5. Lower-middle-left panel: in the office cafeteria or workspace, the woman smiles while talking with coworkers; include desks, monitors, and background people. Add a vertical narration box and a speech bubble from another person. 6. Lower-middle-right panel: close emotional panel of the woman smiling with closed eyes while petting the cat, cheek close to its head. Add a cat sound effect and two narration boxes. 7. Bottom-left wide panel: the woman sits at her desk holding a drink, smiling brightly with eyes closed; coworkers chat in the background. Add a small vertical speech bubble. 8. Bottom-center narrow panel: close-up of the cat’s face with partial hand/petting detail, used as a reflective transition. Add two narrow vertical narration boxes. 9. Bottom-right wide panel: sunset rooftop or riverside overlook from behind; the woman sits beside the cat, both facing a city skyline and dramatic sunburst clouds. Add two narration boxes, one near the sky and one near the lower center. Text content: Include Japanese vertical manga text in the panels. The central theme should be {argument name="story theme" default="a lonely ordinary office worker is saved emotionally by a stray cat and learns to connect with people"}. Use narration similar in tone to: “I am an ordinary office worker anywhere,” “I am not good at talking to people,” “I met one stray cat near the company,” “At first I only watched from a distance,” “Little by little the distance closed,” “Thanks to that child, I became able to look forward,” and “Someday I want to become someone who can give something back.” Visual style: Authentic seinen/shoujo slice-of-life manga, hand-drawn ink, detailed city and office backgrounds, grayscale tones, subtle hatching, cinematic emotional pacing, clean white gutters, no color, no watermark. Keep the woman and cat consistent across all panels, with the cat appearing in exactly 5 panels: eating, sitting alone, being petted, close-up, and sitting beside her at the skyline.

A stylized digital anime illustration of a young East Asian girl with a serious, determined expression, standing with arms crossed over her chest. She has messy, short black hair tied partially into a loose topknot bun with stray strands framing her face, dark expressive eyes, and fair skin. She wears a loose, oversized dark forest-green martial arts kimono robe with light sage inner lining, wide folded cuffs, a black sash (obi) tied at the waist, and striking yellow-gold lightning bolt patterns across the chest. Dynamic ink splatter background in bold black, white, and vivid warm yellow-orange hues, vector art meets comic book illustration style, cel shading, sharp outlines, painterly brush strokes, dramatic lighting, high contrast, vertical portrait --ar 1:2 --v 6.0

Create a single-panel modern manga illustration in a cinematic café interior. The scene shows exactly 2 characters in close-up, waist-up: on the left, a beautiful young woman with long light-brown hair tied half-up with loose strands, large glossy brown eyes, delicate earrings, natural makeup, and a sleeveless ribbed gray turtleneck top; she is smiling and speaking calmly while gesturing with one hand. On the right, a young man with short messy black hair, thick black rectangular glasses, expressive wide eyes, a drop of sweat on his cheek, an open shocked mouth, and a navy hoodie over a white shirt reacts in surprise. Between them, place one large vertical white manga speech bubble with a thick black outline, slightly oval, pointing toward the woman. The speech bubble contains vertical Japanese text in bold black, with the central acronym in bold blue Latin letters. Use the exact visible text arrangement: 「絵は描いてないけど、」 on the right column, blue vertical 「CMOA-I?」 in the middle, and 「漫画を作れるAIを使ったの。」 on the left column. Make the text look like printed manga lettering, sharp and readable, with the blue acronym standing out. Background: softly blurred cozy urban café with large windows, warm hanging lamps, dark walls, a chalkboard menu, plants, and evening light outside. Art style: polished seinen/shoujo manga hybrid, detailed linework, realistic anime faces, soft skin shading, glossy highlights, dramatic rim lighting, warm brown and amber color palette, shallow depth of field, high-resolution digital illustration. Composition: woman occupies the left third, man occupies the right third, speech bubble dominates the center upper area, characters cropped near the shoulders and chest, horizontal 3:2 aspect ratio. Avoid extra characters, extra speech bubbles, logos, watermarks, or unreadable random text.

Goal: Create a cinematic vertical manga page for episode 0 of a near-future domestic sci-fi story, blending photorealistic anime character rendering with clean manga panel borders. Canvas: Vertical 2:3 page, approximately 800 x 1200 px, black gutters and thin white panel borders, warm evening apartment lighting, shallow depth of field, highly detailed semi-realistic anime style. Layout: Use exactly 6 panels arranged as follows: Panel 1 top-left, wide kitchen cooking close-up; Panel 2 top-right, young man on sofa watching TV; Panel 3 middle-left, TV news broadcast screen; Panel 4 middle-right, close-up side profile of the young man; Panel 5 bottom-left, larger panel of the young man reclining with remote; Panel 6 bottom-right, android-like girl turning back toward camera in the kitchen. Characters: The main girl is {argument name="female character description" default="a delicate pale young woman with very long silver-white hair, soft doll-like features, blue glowing android eyes, cream knit sweater, dark apron straps"}. The main boy is {argument name="male character description" default="a slim young man with messy black hair, pale skin, dark hoodie, tired gentle expression"}. The girl appears in panels 1 and 6; the boy appears in panels 2, 4, and 5; the news anchor appears in panel 3. Panel details: Panel 1 shows the silver-haired girl cooking ginger pork in a frying pan in a cozy kitchen, using chopsticks or tongs, steam rising, warm backlight, bottles and plants blurred in the background. Add one speech bubble with Japanese text: {argument name="first speech bubble" default="今日の夕食は\n生姜焼きです"}. Add a small hand-drawn sound effect near the pan: 「ジュ〜」. Panel 2 shows the boy lying on a couch under a blanket, holding a TV remote, looking toward the television; the TV glows blue in the background. Add one speech bubble: 「おー」. Panel 3 is a TV news screen with a serious female news anchor on the left, a mountain forest search-and-rescue image on the right, blue broadcast graphics, and the channel logo “FNN NEWS.” Include a red breaking-news label 「速報」. Add exactly one large white news speech box: 「続いてのニュースです。\n〇〇山で行方不明になっている\n子どもは、現在も発見されておらず\n警察・消防による捜索が続いています。」 Add the lower headline: 「〇〇山で行方不明の子ども、いまだ発見されず」 and small subheadline: 「警察・消防が捜索続ける　悪天候で難航」. Panel 4 shows the boy in side profile, thoughtful, chin resting on hand, warm rim light, TV blur behind him. Add one speech bubble: 「まだ\n見つかって\nないのか……」. Panel 5 shows the boy reclined on the sofa holding the remote, seen from a lower angle with a mug and books softly blurred in the foreground. Add one speech bubble: {argument name="boy concerned line" default="早く見つかると\nいいよな〜"}. Panel 6 shows the silver-haired girl turning from the stove toward the viewer, glowing blue eyes clearly visible, calm expression, steam and kitchen highlights behind her. Add one speech bubble: {argument name="android reply" default="承知しました"}. Visual style: Premium cinematic manga, realistic anime faces, soft skin detail, filmic lighting, warm amber kitchen tones contrasted with cool blue TV-news light, detailed hair strands, glossy eyes, natural domestic setting, polished social-media comic page quality. Constraints: Use exactly 6 panels and exactly 6 speech bubbles plus the news graphics and one sound effect. Keep all Japanese text legible and placed inside rounded white manga speech bubbles or TV captions. Do not add extra characters, extra panels, watermarks, artist signatures, or unrelated text.

Goal: Create a high-energy manga/anime punk music collage featuring two young women in a gritty red-and-black graphic style, as if from a dramatic Japanese music video comic page. Canvas: Wide horizontal 16:9 image, divided into exactly 9 rectangular comic panels arranged in a 3 columns by 3 rows grid, with thick clean white gutters between panels. Use bold black ink outlines, flat cel shading, rough brush textures, red backgrounds, black graffiti strokes, and purple accent typography. Main characters: Show exactly 2 recurring female characters throughout the panels. Character 1 is {argument name="brown-haired woman" default="a young woman with a short light-brown bob haircut, pale skin, wide intense eyes, wearing a plain white T-shirt and loose black pants"}. Character 2 is {argument name="black-haired woman" default="a young woman with long straight black hair and bangs, pale skin, sharp wide eyes, wearing an oversized gray hoodie and loose black pants"}. Their expressions should be manic, deadpan, startled, bored, and confrontational depending on the panel. Keep their faces consistent across all panels. Panel count and layout: Include exactly 9 panels: 1. Top-left panel: close-up of the brown-haired woman lunging toward the viewer with a huge foreshortened hand reaching out, frightened manic expression, red background with black brush marks, purple Japanese text 「あきさん」 near her. 2. Top-middle panel: close-up of the black-haired woman turning over her shoulder in a gray hoodie, surprised eyes, red background, purple handwritten text 「みゆー」 and a simple purple crown doodle. 3. Top-right panel: both women crouching toward the viewer with exaggerated foreshortened hands making framing gestures; the brown-haired woman grimaces and the black-haired woman stares intensely; add handwritten black English text in the corner reading {argument name="corner slogan" default="Music makes us crazy"}. 4. Middle-left panel: black-haired woman in profile looking upward dramatically against a red industrial city skyline and black power-line silhouettes; vertical black label with white Japanese text 「まだ終わらない」. 5. Middle-center panel: both women back-to-back, brown-haired woman facing left and black-haired woman facing right, a purple broken-heart shape behind them; black name labels with white Japanese text 「あきさん」 and 「みゆー」. 6. Middle-right panel: brown-haired woman screaming into a handheld microphone, extreme emotional singing pose; black vertical labels with white Japanese text 「うるさい世界で」 and 「歌ってる」; add a purple crown doodle. 7. Bottom-left panel: black-haired woman curled into her gray hoodie, glaring over her sleeve, purple chains crossing the foreground; purple vertical text 「見て見て見て」. 8. Bottom-middle panel: both women sitting side by side on steps in a red urban alley, bored and exhausted poses with elbows on knees and faces resting on hands; red wall behind them with small handwritten black Japanese text 「ふたりでまだ行けるよ」 and a tiny crown. 9. Bottom-right panel: extreme split close-up of both women’s eyes and partial faces separated by a jagged red lightning slash; include handwritten black Japanese text 「ずっと」 and a small crown. Visual style: Japanese seinen manga meets punk album cover, dramatic fisheye perspective, exaggerated foreshortening, nervous facial expressions, heavy black shadows, red dominant palette, white skin highlights, gray clothing, purple graffiti accents. The art should look hand-inked, chaotic, stylish, and slightly unsettling but polished. Text constraints: Preserve the visible Japanese text exactly as written. Use only the specified text elements, no extra captions, no watermark, no speech bubbles. Make all lettering look hand-painted or manga-style.

A vibrant, cartoonish illustration depicts a {argument name="creature" default="monstrous green creature with bulging eyes and sharp teeth"} driving a souped-up {argument name="car type" default="pink Volkswagen Beetle"}. The car is adorned with flames and the words "BEETLE BOMBER" and "RAT ROD." The creature wears a racing helmet and is shown mid-air, with smoke and flames billowing from the rear tires and exhaust pipes. In the background, a crowd of equally bizarre characters cheers from behind a "DRAG STRIP" sign. Neon signs for "BEETLE JUICE," "COLD BEER," and "NITRO" add to the chaotic, high-energy atmosphere of a drag race. The art style is reminiscent of comic books and graffiti, with bold outlines and saturated colors.

Goal: Create a single black-and-white Japanese yuri manga page about two high school girls awkwardly wanting to walk home together. Canvas: Vertical manga page, approximately 2:3 aspect ratio, clean white page margin, heavy black panel borders, screentone shading, high-detail ink line art, page number “1” centered at the bottom. Visual style: Professional shoujo/seinen manga hybrid, monochrome, delicate facial expressions, glossy hair rendering, soft blush screentones, realistic classroom and hallway perspective, cinematic closeups, dramatic pauses, no color. Use Japanese vertical speech bubbles and narration boxes. Characters: Two teenage schoolgirls in dark blazers, white shirts, neckties, and pleated skirts. Girl A has {argument name="black haired girl name" default="the quiet black-haired girl"}, very long straight black hair with blunt bangs, shy downcast eyes, subtle blush, reserved posture. Girl B has {argument name="blonde girl name" default="the short-haired blonde girl"}, short layered light hair with dark roots, sharper eyes, slightly princely/tomboyish aura, a school bag, embarrassed but trying to act casual. Layout: Use exactly 6 manga panels arranged from top to bottom: 1. Wide establishing panel across the top: an empty classroom after school with desks, chalkboard reading 「また あした」 with a small flower doodle, large windows and curtains. Girl A sits at a desk with her bag while Girl B stands beside her leaning forward. Add a tall vertical narration box on the right reading 「二人きりになると、うまく息ができない。」 2. Middle-left closeup panel: Girl A’s face looking down, intensely blushing, long black hair framing her cheeks. Speech bubble reads 「……うん。」 3. Middle-right closeup panel: Girl B smiling nervously with flushed cheeks. Speech bubble reads 「今日も、一緒に帰る？」 4. Lower-middle-left panel: Girl B looking down sadly or uncertainly, eyes softened, with a silent speech bubble reading 「……」 5. Lower-middle-right detail panel: closeup of Girl A’s hands gripping the handles of a school bag tightly, with small sound effect 「ぎゅっ」 and two vertical narration boxes reading 「たった二文字。」 and 「本当は、ずっと待っていたのに。」 6. Wide bottom panel: school corridor with windows and perspective depth. The two girls stand side by side; Girl B looks away shyly while holding her bag strap, Girl A looks surprised and blushing. Girl B’s speech bubble reads 「ねえ。私と帰るの、嫌じゃない？」 and Girl A’s small rectangular reaction box reads 「え？」 Text constraints: Keep all Japanese text exactly as written, vertical where appropriate. Do not add extra dialogue, sound effects, logos, watermarks, or modern digital UI elements. The scene should feel tender, quiet, romantic, and emotionally restrained.

Create a dramatic monochrome fantasy illustration of {argument name="creature name" default="Orthrus"}, a massive two-headed hellhound charging straight toward the viewer through a dense ancient forest. The creature has exactly 2 snarling canine heads: the left head faces forward with jaws wide open, long fangs exposed, tongue visible, ears raised and spiked fur bristling; the right head is slightly behind and turned to the right, also roaring with sharp teeth bared. Give the beast a muscular wolf-like body, thick shaggy black-gray fur, huge clawed front paws lunging over fallen logs, a curling tail visible on the left, and heavy metal collars, rings, and chain links around both necks. Set the scene in a chaotic woodland with tall tree trunks, tangled branches, leaf litter, splintered deadwood, rocks, and dust or debris kicked up under its paws. Use an intense low frontal perspective, close crop, dynamic motion, and a sense of imminent attack. Render in gritty black-and-white ink, charcoal, and graphite crosshatching, with dense scratchy linework, high contrast, rough paper texture, expressive dark shadows, no color, no text, no border, no watermark. Emphasize ferocity, mythological horror, and hand-drawn dark fantasy concept art in a square composition.

一个女孩说着“{argument name="girl's dialogue" default="嘿，嘿，听我说！"}”，以及一个位于男孩上方的方形气泡，显示他的内心想法“{argument name="boy's inner thought" default="我就假装没听见吧"}”。以 {argument name="aspect ratio" default="竖向 (4:5)"} 的纵横比进行渲染。

Create an ultra-detailed dark fantasy gothic illustration of {argument name="character name" default="an undead vampire king"} enthroned inside a ruined cathedral throne room. The central figure is a pale, corpse-like monarch with long flowing {argument name="hair color" default="white hair"}, glowing red eyes, blood running from the mouth, and a jagged black crown. He wears ornate blackened armor covered in skulls, spikes, chains, red jewels, and torn crimson-black royal robes. He sits arrogantly on a skull-encrusted stone throne, one hand resting on the armrest and the other gripping a long ceremonial sword with a skull pommel, gothic crossguard, red gem, and pale blade. Surround him with exactly 9 clearly visible supernatural figures or creatures: 1 large translucent ghostly warrior hovering behind him with a sword, 1 smaller floating ghost near the moonlit archway, 4 misty human spirits on the stairs and balcony in the midground, 2 armored spectral soldiers at the lower right holding weapons, and 1 snarling ghost wolf in the lower left foreground. Set the scene in a vast black stone cathedral with broken pointed arches, towering columns, candles, chains, cracked steps, skulls, and tattered red banners. Include exactly 3 readable banner or stone inscriptions: left hanging banner says {argument name="left banner text" default="BLOOD BEYOND DEATH"}, right hanging banner says {argument name="right banner text" default="KINGS DIE BUT LEGENDS HAUNT FOREVER"}, and the stone block near the lower left says {argument name="stone inscription" default="SWORDS REMEMBER WHAT MEN FORGET"}. Through the huge archway, show a distant spired gothic castle, an arched bridge, flying bats, storm clouds, cold mist, and a huge crimson blood moon. Use cinematic low-angle composition, hyperreal digital painting, intricate concept-art detail, cold blue-gray fog contrasted with deep crimson accents, dramatic moonlight, candle glow, volumetric mist, sharp armor highlights, macabre regal atmosphere, no modern objects, no cartoon style.

Goal: Create a vertical retro Teacher's Day tribute poster in a limited navy blue, warm orange-red, and cream palette, with a heroic educational illustration and bold typography. Canvas: Portrait poster, 2:3 aspect ratio, off-white paper background with subtle grain, aged print texture, risograph/screenprint look, slight ink misregistration and distressed halftone shading. Layout: Place a huge headline across the top reading {argument name="headline text" default="TEACHER’S DAY"} in tall condensed vintage serif capitals, navy blue with orange-red offset shadows and rough ink edges. Under it, center a large orange-red Chinese line reading {argument name="Chinese message" default="老师，辛苦了"}. Beneath that, add a small navy serif subtitle reading {argument name="subtitle text" default="Thanks for lighting the way"}. The lower two-thirds contains the illustration: a monumental female teacher rising behind oversized books, looking upward toward the right, while a tiny student stands at the bottom right for scale. Subject details: The main teacher is one adult woman with short wavy dark blue hair, round glasses, a calm proud upward gaze, pale cream skin, and a dark navy robe or cardigan over a light blouse. She holds exactly 2 books against her chest: one orange-red cover book in front and one cream/white book with navy outlines behind it. In the foreground, show exactly 3 oversized stacked books spanning the lower half: top book is mostly cream pages with navy and orange-red edges, middle book has an orange-red cover, bottom book has cream pages and a dark navy cover/edge. Add exactly 1 tiny student at the bottom right, wearing a navy school uniform and carrying an orange-red backpack, facing the teacher. Add stylized plant leaves behind the book stack, using navy and orange-red shapes, as decorative background elements. Visual style: Inspirational editorial poster illustration, simplified anatomy, elegant linework, high-contrast block colors, rough pencil-like crosshatching in the navy areas, warm nostalgic Chinese poster influence, clean centered composition, no photorealism. Constraints: Use exactly 3 visible text lines: the large English headline, the Chinese message, and the small English subtitle. Keep the typography centered and legible. Do not add logos, watermarks, extra people, classroom objects, or additional text.

Goal: Create a wide monochrome comparison sheet showing five nearly identical Japanese manga pages placed side by side, like an image-generation quality comparison. Each page should depict the same cute fox-eared girl food-shopping scene, with tiny labels centered above each page for different quality settings. Canvas: Extremely wide horizontal canvas, about 5:1 aspect ratio, white background, crisp black-and-white manga printing. Place exactly 5 complete manga pages in a row with narrow white gaps between them. Each page is vertical and identical in overall composition, using clean screentones, thin panel borders, speed lines, and detailed manga backgrounds. Layout per page: Each of the 5 pages contains exactly 8 panels arranged in 4 tiers: 1 wide top panel, 2 panels in the second tier, 3 panels in the third tier, and 2 panels in the bottom tier. Keep the repeated pages almost the same, with only subtle quality/detail differences between them. Page labels: Add 5 tiny English labels above the pages, one per page: “quality low”, “quality medium”, “quality high”, “quality ultra”, and “quality max”. Story and panel details for every page: 1. Top wide panel: A cheerful anime fox girl with large fox ears, long light hair, and a casual hoodie smiles brightly in a cozy room with a window, curtains, plants, and a round lamp behind her. She raises one hand in excitement. Include a vertical oval speech bubble containing {argument name="top speech bubble" default="今日はお弁当を買いに行こう"}. 2. Second-tier left panel: Close-up of prepared boxed lunches or bento trays in a display case, drawn with detailed food textures and packaging. 3. Second-tier right panel: Exterior of a modern convenience store or deli storefront with glass windows, a signboard, sidewalk, and small background figures. Add two vertical speech bubbles with small Japanese manga text. 4. Third-tier left panel: Rear view of two or three customers standing in line inside the shop; the fox girl is visible among them holding a basket. 5. Third-tier middle narrow panel: A small dialogue/reaction panel with a simple speech bubble, suggesting hesitation or ordering. 6. Third-tier right panel: The fox girl at a counter or table looking at a bento box, smiling and holding chopsticks or a tray, with another speech bubble. 7. Bottom-tier large left panel: Dramatic close-up of the fox girl shocked and overwhelmed by delicious food or the amount she bought; wide eyes, open mouth, blush, speed lines radiating behind her, and a basket or bento boxes in the foreground. Include a bold speech bubble with {argument name="reaction speech bubble" default="やべぇ ついつい買ってるじゃん"}. 8. Bottom-tier small right panel: A simplified stunned reaction portrait of the fox girl facing forward, blank expression, large ears, and a tiny vertical bubble with ellipsis. Character details: The main character is {argument name="character description" default="a young anime fox girl with fluffy upright fox ears, long pale hair with bangs, bright eyes, and a playful expression"}. Her outfit is {argument name="outfit" default="a loose casual hoodie with a simple collar"}. Use a cute modern manga style, expressive face acting, clean linework, and soft grayscale screentone shading. Visual style: Black-and-white manga page art, high detail, sharp ink lines, halftone shading, traditional Japanese comic paneling, realistic convenience-store backgrounds, readable panel borders, no color except grayscale. The five pages should feel like a side-by-side technical comparison, not separate unrelated illustrations. Constraints: Use exactly 5 repeated pages and exactly 8 panels per page, for a total of 40 visible comic panels. Do not add extra characters beyond small background shoppers and staff. Keep all nonessential dialogue as tiny Japanese-style manga lettering, and avoid watermarks or logos.

gritty {argument name="style" default="horror-punk"} collage in newsprint, torn paper, xerox grain, and {argument name="color" default="blood-red"} paint.

Create a premium cinematic character bible sheet for {argument name="character name" default="BABY"}. Use uploaded character sheet as strict visual reference. Do not change his appearance. LEFT SIDE: Massive bold brushstroke text {argument name="character name" default="BABY"} top left, {argument name="accent color" default="deep red"}. Small text: {argument name="subtitle" default="THE DRIVER / BABY DRIVER"}.