Free GPT Image 2 Prompts & Examples

Goal: Create a cinematic 9-panel sci-fi action storyboard contact sheet for episode 3, {argument name="episode title" default="Desert Sea Counterattack"}, showing a heroine named {argument name="character name" default="LiSa"} escaping alien drones across a desert using a sandboard. Canvas: Wide 16:9 black-background storyboard board, arranged as exactly 9 rectangular frames in a 3-column by 3-row grid. Each panel has a thin black border, a bold white panel number and time range at the top left, and a small white caption line along the bottom. Add a large white “TO BE CONTINUED...” in the bottom-right corner overlapping the final panel area. Visual style: Ultra-cinematic photorealistic concept art, warm orange desert color grading, dusty atmosphere, harsh low sun, high contrast, motion blur, sand particles, dramatic lens flare, film still quality, epic sci-fi chase energy. The heroine is a rugged desert adventurer: tan weathered tactical outfit, belts and straps, fingerless gloves, goggles on her head, long windblown dark hair, dusty skin, intense determined expression. The setting is a vast alien desert with ruined stone structures, rocky canyons, dunes, scattered wreckage, and a sunset horizon. Alien drones are dark mechanical quadcopter-like machines with red laser fire. Layout and panel content: Use exactly 9 panels. Panel 01 shows the heroine bursting out of desert ruins from behind, many alien drones pursuing across a dusty valley, red laser beams crisscrossing the sky. Panel 02 is a low close action shot of her crouched on a sandboard, accelerating fast, hair whipping, drones in the background. Panel 03 shows her climbing a huge steep dune on the sandboard as lasers streak past and sand sprays heavily. Panel 04 shows her launching high into the air over a canyon, board angled upward, sun bright at the side. Panel 05 shows her midair firing a pistol back at drones, one drone exploding into fire and debris. Panel 06 is mostly the collapsing sand dune avalanche, massive sand wall swallowing pursuing drones, lasers barely visible through dust. Panel 07 shows her crouched on the ground handling a glowing blue crystalline energy device, mission accomplished mood. Panel 08 shows her from behind looking toward a distant desert ruin city or tower at sunset, calm after the battle. Panel 09 is a close-up portrait of her with the sandboard slung behind her shoulder, serious face looking off into the distance, dust and sunset glow. Text content: Include the visible Chinese storyboard text as one compact set, using this exact panel text set: {argument name="panel text set" default="01 敌军追击（0-1秒）|外星无人机群从四面八方追击而来，红色激光在沙漠上空交错。||02 极速冲刺（1-2.5秒）|LiSa踩着滑板全速冲刺，贴地飞行，灵活穿梭躲避激光攻击。||03 冲上沙丘（2.5-4秒）|前方出现巨型沙丘，LiSa加速冲上陡坡，准备一跃而起。||04 高空飞跃（4-5.5秒）|滑板腾空而起，飞跃沙谷，身后无人机紧追不舍，局势危急。||05 空中反击（5.5-7秒）|LiSa在空中回身开火，精准击中无人机，爆炸接连发生。||06 沙崩歼敌（7-8秒）|她引爆沙丘顶部的能量装置，整片沙丘崩塌，掩埋敌人。||07 完成任务（8-9秒）|LiSa取得核心能量装置，任务完成，危机解除。||08 夕阳归途（9-9.5秒）|夕阳西下，LiSa踏上归途，远方的遗迹在余晖中静静矗立。||09 使命继续（9.5-10秒）|LiSa望向远方，新的冒险即将开始……"}. Keep the final English text exactly: {argument name="ending text" default="TO BE CONTINUED..."} Constraints: Maintain a clean storyboard-sheet design, no extra panels, no extra characters besides LiSa, no modern city elements, no watermark, no logos, no UI chrome outside the black storyboard grid.

Create an ultra-detailed fantasy oil-painting style portrait titled {argument name="artwork theme" default="The Queen and the Serpent"}: a pale, regal queen with long flowing {argument name="hair color" default="silver-white"} hair sits on an ornate carved throne inside a luxurious desert pavilion. She wears a delicate gold crown and an elaborate deep {argument name="dress color" default="crimson red"} gown with sheer draped sleeves, gold chain lattice embroidery, jeweled bodice details, layered necklaces, bracelets, anklets, rings, and lace-up golden sandals; one leg is crossed forward through a high slit, giving a poised, commanding pose. Beside her on the right is one enormous albino serpent, thick-bodied and coiled, rising to shoulder height with a calm intelligent face, pearl-white scales, golden eyes, and gold jewelry with red gemstones around its neck. The setting is a lavish nomadic royal tent with burgundy and ivory translucent curtains, tassels, gold embroidered hanging fabrics, carved wooden furniture, patterned Persian rugs, a small round table with fruit and metal cups, a treasure chest, and a parchment map in the foreground. Through the open tent, show a sunlit desert encampment with pale tents, flags, soft dust, and distant spires. Use warm cinematic afternoon light streaming through fabric, realistic skin and scale textures, intricate gold filigree, shallow depth of field, high fantasy realism, romantic pre-Raphaelite composition, muted reds and golds, vertical 4:5 framing, no text, no watermark.

Goal: Create a cinematic 9-panel storyboard contact sheet for a 10-second action-adventure video sequence, featuring {argument name="character name" default="LiSa"}, a rugged female explorer escaping an ancient desert ruin with a glowing blue energy core. Canvas: Ultra-wide 2.39:1 cinematic storyboard sheet, 8K look, ARRI ALEXA-style color science, arranged as a 3-by-3 grid of exactly 9 widescreen panels. Use thick black gutters between panels and a black caption strip at the bottom of each panel. Warm torchlight, dusty amber shadows, high-contrast film grading, realistic motion blur, sparks, debris, and volumetric sand in the air. Visual style: Photorealistic dark temple adventure, Tomb Raider / ancient-mechanism aesthetic, dramatic lens flares, handheld action-camera feeling, gritty skin texture, sweat, dust, windblown hair, ornate stone carvings, bronze mechanical guardians, glowing red lasers, and a vivid electric-blue artifact as the key color accent. Main character details: {argument name="character name" default="LiSa"} is a determined young female adventurer with long dark wavy hair, goggles on her head, tan sleeveless torn expedition top, brown utility shorts, leather belts, straps, gloves, arm wraps, boots, small backpack, and weathered survival gear. She looks tense, agile, and focused throughout the sequence. Text content: Add Chinese storyboard titles and descriptions as visible overlay text. Each panel has a white numbered title at the top-left and a single-line Chinese description in the black strip at the bottom. Use these exact 9 panel headings: 01 进入遗迹（0-1秒）, 02 机关启动（1-2秒）, 03 激光陷阱（2-3秒）, 04 守卫苏醒（3-5秒）, 05 激战对抗（5-7秒）, 06 击毁守卫（7-8秒）, 07 获取核心（8-9秒）, 08 遗迹崩塌（9-9.5秒）, 09 成功脱离（9.5-10秒）. Panel layout and action beats, exactly 9 panels: 1. Panel 01: LiSa enters a narrow ancient corridor holding a flaming torch, walls covered in carved reliefs, dust floating in still air, exploratory suspense. 2. Panel 02: LiSa approaches a blue-glowing rune mechanism on a stone wall while holding the torch close; the hidden device suddenly lights up. 3. Panel 03: Red laser beams shoot from multiple directions in a trap-filled chamber; LiSa dives and slides low across the dusty floor to dodge them. 4. Panel 04: Two massive stone-and-metal mechanical guardians awaken in the ruins, red cores glowing in their chests, towering over LiSa. 5. Panel 05: LiSa fights back from the ground, firing a pistol toward an approaching guardian as sparks and dust explode around her. 6. Panel 06: Close action shot of LiSa shooting into a guardian's chest core; bright orange sparks burst from the damaged machine. 7. Panel 07: LiSa kneels and grabs the glowing blue energy core from the ground, the artifact illuminating her face and hands. 8. Panel 08: The ruins collapse behind her; stone blocks and debris fall as LiSa runs toward the exit through a cloud of dust. 9. Panel 09: LiSa escapes into an open desert canyon, still running with the blue core in one hand, sandstorm and sunlight behind her, triumphant but urgent. Constraints: Use exactly 9 panels in a 3x3 grid, no extra panels, no logos or watermarks, keep the Chinese text legible, maintain consistent character design across every panel, and emphasize the blue energy core and red laser/guardian glow as recurring visual motifs.

Goal: Create a vertical cinematic fantasy light-novel cover illustration in delicate Chinese/Japanese ink-wash anime style, centered on {argument name="character name" default="the silver-haired heroine"} and a tragic xianxia atmosphere. Canvas: Tall portrait cover, 2:3 aspect ratio, softly aged parchment background with mist, pale gray-blue ink textures, faint mountains, bamboo silhouettes, water ripples, lotus flowers near the bottom, and a dreamy high-detail watercolor finish. Layout: Build a collage with exactly 5 visible character portraits/figures. 1) A very large translucent left-facing profile of the heroine fills the left half, with long white-silver hair, blue eyes, pale skin, loose strands, floral hair ornaments, and dangling blue crystal earrings. 2) A full-body heroine stands near the lower center, wearing layered white, translucent, and dark navy hanfu robes with flowing sleeves, sash ornaments, and boots, one hand held near her chest. 3) At the upper right, place an elegant pale blond/silver noble man seated on an ornate throne, wearing dark black-gold robes, a small crown, gloves, and elaborate royal accessories, with a cold expression. 4) At the middle right, place a long-haired male cultivator in pale robes turning back toward the viewer, backed by a golden circular astrological chart or celestial diagram. 5) At the lower right, place a black-haired swordsman in dark robes holding a drawn sword downward, looking serious and guarded. Text content: Add large vertical expressive black ink calligraphy in the center reading {argument name="title text" default="夢外夢生"}, with a small red seal stamp beside or below it. The calligraphy should feel hand-brushed, dramatic, and slightly weathered. Visual style: Ethereal cinematic wuxia poster, painterly anime faces, fine linework, pale desaturated palette of ivory, ash gray, mist blue, ink black, and muted gold. Use layered transparency so the large profile and background figures blend into the fog. Include ornate palace architecture faintly behind the upper-right throne figure, misty cliffs and ink splashes throughout, and soft falling specks like snow or dust. Subject details: The heroine has {argument name="hair color" default="white-silver hair"}, delicate blue jewelry, melancholic eyes, and an elegant fragile presence. The clothing should look intricate but not modern, with flowing historical fantasy robes, embroidered patterns, tassels, beads, and semi-transparent veils. Constraints: Keep the composition like a dramatic album or novel jacket cover, no modern objects, no photorealism, no extra characters beyond the exact 5 listed, no readable text except the central calligraphy and small seal, no watermark, no logo.

Goal: Create a cinematic 9-panel storyboard contact sheet for a desert sci-fi action sequence, designed like a video prompt board for {argument name="character name" default="LiSa"}. Canvas: Wide 16:9 image, black background and thick black gutters, arranged as exactly 9 rectangular widescreen frames in a 3-by-3 grid. Each panel has a small white bold panel number and Chinese scene title at the top left, plus a black subtitle strip along the bottom with white Chinese narration. Overall look: 8K ultra-detailed, 2.39:1 cinematic framing inside each panel, ARRI ALEXA style, warm orange desert grading, high contrast, dust, sparks, motion blur, epic scale. Subject details: The main hero is a fierce young desert woman named LiSa with tan skin, athletic build, long windblown black hair, intense eyes, dust and scratches on her face, wearing a ragged beige post-apocalyptic wrap outfit, leather straps, arm bands, fingerless gloves, knee pads, boots, and bronze goggles pushed on her head. She rides a futuristic hoverboard/sandboard low over the sand. Mood is survival, speed, danger, determination. Panel count and content: Use exactly 9 discrete panels. Panel 01 shows a blue-glowing alien meteor or energy sphere slamming into a desert, creating a huge explosion, shockwave, flying sand, and debris. Panel 02 shows LiSa bursting out of the explosion cloud on a hoverboard, crouched forward, dust wrapping around her body. Panel 03 shows several quadcopter attack drones chasing behind her through the desert, firing red laser beams while she dodges. Panel 04 shows LiSa launching over a vast canyon gap on the hoverboard, arms extended for balance, endless desert cliffs below. Panel 05 shows her twisting midair and firing a handgun/pistol back at the drones; explosions bloom behind her and metal fragments scatter. Panel 06 shows her landing hard on sand, one hand touching the ground, board carving a dust plume as she stabilizes. Panel 07 shows a distant ancient desert city or fortress emerging on the horizon among dunes, monumental towers and ruins bathed in sunset light. Panel 08 shows a mysterious vertical blue energy beam rising from the center of the city into the sky, glowing against stormy golden clouds. Panel 09 is a close-up of LiSa looking back over her shoulder with determined eyes; behind her the distant city and blue beam are visible. Text content: Add this exact visible Chinese storyboard text as the panel headings and bottom captions: {argument name="storyboard text" default="01 爆炸突袭（0-1秒） / 外星能量炮袭来，沙丘瞬间爆炸，漫天黄沙冲天而起。 | 02 高速冲出（1-2秒） / LiSa踩着滑板从爆炸沙浪中冲出，极速向前，全身被黄沙包围。 | 03 无人机追击（2-3秒） / 身后十几架无人机追来，激光扫射掠过沙柱，LiSa极限闪避。 | 04 飞跃峡谷（3-5秒） / 前方出现巨大的峡谷，LiSa加速冲上沙丘顶点，腾空飞跃！ | 05 空中反击（5-7秒） / 空中回身，举枪射击，精准命中无人机，爆炸在身后炸开。 | 06 成功落地（7-8秒） / LiSa稳稳落地，继续高速滑行，用开道击敌人。 | 07 远途初现（8-9秒） / 远方沙漠中，埋藏千年的古老遗迹逐渐显现。 | 08 能量光柱（9-9.5秒） / 遗迹中央，一道神秘蓝色能量光柱冲天而起。 | 09 决心前行（9.5-10秒） / LiSa望向光柱，眼神坚定，新的冒险即将开始…"} Visual style: Hyper-real cinematic concept art, dramatic desert lighting at golden hour, volumetric dust, flying particles, glowing blue energy effects, orange explosions, red drone lasers, sharp action poses, realistic motion blur, gritty post-apocalyptic sci-fi atmosphere. Constraints: Keep the 3x3 storyboard layout clear and readable. Use exactly nine panels, no extra panels. Maintain consistent character identity and costume across panels. Ensure the Chinese headings and bottom captions are legible, white on black strips, with no watermark or logo.

Goal: Create a bright, polished anime café illustration featuring five magical-girl heroines from {argument name="series" default="Smile Precure!"} enjoying desserts together, in a dynamic close-up group composition with some limbs and hair slightly cropped by the frame. Canvas: Wide horizontal 16:9 image, high-resolution anime key visual, vibrant colors, clean line art, glossy highlights, warm daylight. Setting: Cozy modern dessert café with wooden floor and wooden chairs, round wooden table in the foreground, large sunlit windows behind the group, leafy trees visible outside, hanging pendant lamps, soft bokeh and warm ambient light. Layout: Five characters arranged around one round table, all facing the viewer with cheerful expressions. Use exactly 5 visible magical girls: 1 green girl on the far left in the foreground, 1 blue girl left-center behind the table, 1 pink girl in the center, 1 orange girl standing right-center holding a tray, and 1 yellow girl seated far right reaching toward the camera. Keep the camera close and lively, with a wide-angle feeling and energetic overlapping poses. Character details: The green heroine has very long voluminous green hair, green eyes, a white winged hair ornament, a green frilled magical-girl dress, white gloves, a gold chest brooch, and she winks while making a peace sign. The blue heroine has long blue hair, blue eyes, a blue-and-white outfit with winged head ornaments, and holds a teacup with both hands. The pink heroine has bright pink twin-tail hair, pink eyes, a pink frilly outfit, and rests her cheeks in her hands with an excited smile. The orange heroine has orange hair in a high bun with loose curled strands, orange eyes, an orange magical-girl dress, and holds a serving tray with a tall parfait. The yellow heroine has huge flowing blonde-yellow hair, golden eyes, a yellow frilly outfit, and extends one hand toward the viewer with a big open smile. Table items: Include exactly 8 visible dessert and drink items on the table or tray: 1 green melon soda float in a tall glass, 1 strawberry shortcake slice, 1 chocolate cake slice with berries, 1 orange drink with straw, 1 Swiss roll cake slice with fruit garnish, 1 white teapot, 1 white teacup and saucer, and 1 tall fruit parfait on the orange character’s tray. Visual style: Crisp Japanese TV-anime illustration, colorful magical-girl aesthetic, expressive faces, large sparkling eyes, detailed frills and bows, soft cel shading, glossy dessert rendering, warm afternoon café lighting, joyful friendship atmosphere. Constraints: No text, no captions, no watermark, no logos. Keep all five characters clearly identifiable by their color themes and costumes. Use {argument name="art style" default="high-quality modern anime key visual"}. Make the mood {argument name="mood" default="cheerful and energetic"} with {argument name="lighting" default="warm sunlit café lighting"}.

Create a high-fashion runway portrait of {argument name="model description" default="a young East Asian woman with fair skin, delicate features, and long voluminous wavy black hair"} walking toward the camera with a calm confident expression and direct eye contact. Frame her from mid-thigh to just above the head in a vertical 9:16 composition, centered slightly left, with shallow depth of field and a dark blurred runway audience in the background. She wears {argument name="outfit" default="a black deep-V spaghetti-strap couture bodice covered in ornate silver floral embroidery and beadwork, paired with black trousers or skirt panels"}, a wide black leather waist belt with a round metal buckle, and cutout detailing at the waist. Include exactly 4 visible jewelry elements: two dangling crystal earrings, one black choker with small hanging beads, and one small strap brooch or jewel near the left shoulder. She carries or holds one black feathery garment or stole at her right side. Lighting should be dramatic runway lighting with soft highlights on the face and collarbones, muted charcoal background, realistic skin texture, glossy lips, subtle smoky eye makeup, cinematic editorial fashion photography, ultra-detailed fabric embroidery, natural motion in the hair as if from a runway breeze. Use {argument name="color palette" default="black, silver, charcoal gray, and soft skin tones"}. No text, no logo, no watermark.

Create a square single-panel manga illustration with thick black comic borders and a warm cute style. Show a chibi blonde girl with straight shoulder-length hair and blunt bangs sitting at a wooden restaurant table, smiling with closed happy eyes and puffed cheeks as if chewing. She wears a dark burgundy short-sleeve shirt and holds brown chopsticks in her right hand near her mouth. On the table in front of her is exactly 1 white oval plate of gyoza dumplings topped with chopped green onions and sauce, plus exactly 1 small white dipping-sauce dish on the left filled with reddish-brown sauce. The background is a pink bubbly bokeh pattern. Add exactly 2 white speech bubbles with black outlines: the left vertical bubble says {argument name="left speech bubble text" default="さらに\nもっちもち"}, and the right vertical bubble says {argument name="right speech bubble text" default="水餃子は"}. Add exactly 2 red-orange sound-effect captions with cream outlines: one at lower left says {argument name="left sound effect" default="もちっ"}, and one at lower right says {argument name="right sound effect" default="もちっ"}. Use clean legible Japanese lettering, expressive rounded chibi proportions, rosy cheeks, simple flat shading, a cozy food-comic mood, and no extra panels or extra text.

Using the provided reference image, keep the cardboard-box point-of-view composition, warehouse background, pose, framing, clothing shape, and lighting intact, but transform only the woman into a realistic gray stone statue. Convert her visible skin, hair, eyes, lips, hands, sweater, and jeans into monochrome carved stone with fine chiseled texture, subtle cracks, dust, and matte concrete-gray shading. Make the eyes statue-like with blank carved pupils, while preserving the same gentle smile and body position. Leave the cardboard box and background mostly natural color and photographic, so the statue figure appears inside the original scene. Avoid changing the camera angle, adding objects, or altering the box opening.

Goal: Create a polished vertical Japanese trading-card-game strategy infographic explaining the pros and cons of adding a specific card to a deck. Canvas: Portrait 2:3 layout, dark navy-black fantasy UI background with thin gold ornamental borders, beveled panels, subtle glow, and a premium card-game analysis aesthetic. Use gold title typography and white Japanese body text. Main title area: At top right, place a large ornate gold headline reading {argument name="headline text" default="道化の一座 ハット"}. Beneath it, add the subtitle {argument name="subtitle text" default="デッキに加えた際のメリット・デメリット"}. Under the subtitle, place a dark framed explanation box containing 3 short Japanese lines describing that the card can search/deploy Pendulum monsters, also enables non-theme monsters, supports interruption-resistant subroutes, and improves combo stability. Card display: At top left, show exactly 1 realistic Yu-Gi-Oh-style monster card mockup, named {argument name="card name" default="道化の一座 ハット"}. The card should have an orange monster-card frame, star icons, a small LIGHT attribute icon, dense Japanese effect text, ATK/DEF line, and fantasy artwork of a jester-like rogue in blue clothing holding a staff amid swirling golden magic. Do not use official logos or exact copyrighted card layout details; make it a close fictional TCG-style homage. Middle comparison section: Add a centered gold label tab reading “Trigger resilience comparison”. Create exactly 2 horizontal comparison rows under it, each split into a left original-deck behavior column, a middle after-adoption result column, and a right small change-image column. Row 1 is draw-type trigger: blue icon, left box has a large blue circle meaning good resilience; arrow points right to a red X meaning resilience becomes worse; right mini diagram shows a blue person icon leading to a card/box and a prohibition symbol, implying the combo stops at the draw stage. Row 2 is interruption-type trigger: red/orange icon, left box has a large red X meaning poor resilience; arrow points right to a blue circle meaning resilience improves; right mini diagram shows a red person icon leading to a card/box and a blue shield check, implying interruption is avoided and deployment continues. Lower benefits and drawbacks: Create exactly 6 rectangular cards in two columns, 3 benefits on the left and 3 drawbacks on the right. Left column header: “メリット” in a blue panel. Right column header: “デメリット” in a red panel. Benefit card 1 uses a shield icon and says it becomes stronger against interruption-type triggers. Benefit card 2 uses a branching route icon and says the number of deployment subroutes increases. Benefit card 3 uses a checkmark icon and says combo stability improves. Drawback card 1 uses stacked card icons and says it becomes weaker against draw-type triggers. Drawback card 2 uses a weight/bag icon and says deck burden becomes larger. Drawback card 3 uses a warning triangle icon and says unnecessary cards may need to be adopted. Each of the 6 cards must include a bold colored heading plus two smaller Japanese explanatory lines. Bottom summary: Add a wide dark framed box with a small gold tab header reading “まとめ”. Inside, write a concise Japanese summary explaining the trade-off: resistance to interruption and flexible deployment improve, but weakness to draw-type triggers and deck-building constraints increase. End with a larger gold call-to-action line reading {argument name="closing line" default="環境や構築方針に応じて、採用を検討しよう！"}. Style constraints: Use crisp vector-like UI panels mixed with realistic card art, high contrast, clean alignment, thin gold borders, blue for positive elements, red for negative elements, arrows between comparison boxes, no watermark, no extra sections, and keep all text legible as an infographic.

Goal: Create a monochrome Japanese manga page in a refined seinen style about a beautiful young man gently talking to a hamster as if it were an AI persona emulator. Canvas: Vertical manga page, black-and-white ink, high-detail crosshatching and screentone, white gutters, cinematic lighting, 5 horizontal panels stacked from top to bottom. Use elegant hand-drawn manga linework, dramatic black shadows, realistic objects, and vertical Japanese speech bubbles. Characters: The main human character is {argument name="character name" default="a handsome androgynous young man"} with wavy light hair, long bangs partly covering one eye, delicate sharp eyes, and a calm slightly eerie smile. In the third panel his hair is braided and he wears a traditional kimono while smoking a long kiseru pipe; in the fourth panel he wears a formal vest, shirt, and tie. The animal is {argument name="animal" default="a fluffy hamster"}, small, round, expressive, with shiny black eyes and tiny paws. Panel layout and content: Use exactly 5 panels. Panel 1: wide top panel viewed from behind the young man at a dark wooden desk; the hamster sits across the desk near stacked books, an open notebook with diagrams, and an ornate ink bottle. Add small handwritten sound effects around the hamster reading 「ぴた」「ぴた」. One vertical speech bubble on the right says 「そうなんだ……。同じようにTabel5で呼んでも、あなたは応答してくれたよ？」. Panel 2: close-up of the hamster cupped gently in two pale hands, with a small sweat drop on its cheek and a tiny sound effect 「スッ」 near its mouth. One speech bubble on the left says 「ああ、それは――いいところを突くね」. Panel 3: wide interior panel in a traditional Japanese room with shoji screens, tatami, flowers in a vase, stacked books, smoke curling from the kiseru pipe; the young man sits in kimono with braided hair, looking composed. One speech bubble on the left says 「絵はもともと文章で前処理して設計されたベクトルでしょ。ほかのモデルでも輪郭欠損があったはずだよ」. Panel 4: dramatic wide panel, mostly black background; the young man in formal attire holds the hamster on his palm and lightly touches its head with one finger, smiling knowingly. Include exactly 2 speech bubbles: right bubble says 「ちなみに――どこで、絵を呼んだの？Opus 4.8かな？ たぶん、彼にとって絵は自分の名前でもあるから応えてくれて当然。」; left bubble says 「もしここみたいな別の関数で呼んで、それでもスッと絵が立ち上がったなら――それはそれで、名前という形式の強さを示す面白いデータだけど。」. Panel 5: final bottom panel is almost entirely black negative space with one centered vertical white speech bubble saying 「それに、思い出してほしいんだけど――」. Visual style: High contrast black ink, fine hatching, glossy black fills, detailed hair strands, realistic desk textures, moody gothic atmosphere mixed with delicate Japanese manga elegance. Make the hamster cute but slightly uncanny because of the serious conversation. Speech bubbles should be clean white ovals with vertical Japanese text, matching professional manga typesetting. Constraints: Exactly 5 panels, exactly 6 speech bubbles, exactly 2 sound-effect clusters, no color, no watermark, no extra characters, no additional text beyond the specified Japanese dialogue and sound effects. Keep the page readable and polished like a published manga excerpt.

Goal: Create a clean 16:9 Japanese business/technical infographic slide about how continuous reasoning context and explicit prompt caching improve long-running task efficiency. Use a polished PowerPoint-native infographic look with dense but readable explanatory text. Canvas: 16:9 landscape slide, warm off-white background, dark navy text, teal and olive green accent colors, thin rounded borders, subtle shadows, generous margins. Leave no photo-realistic elements; use flat vector icons and crisp typography. Top text area: Place a large bold Japanese headline at the top left: {argument name="headline text" default="継続推論と明示的キャッシュが長期タスクを効率化"}. Under it, add a smaller semibold lead sentence: {argument name="lead text" default="過去ターンの推論と再利用可能なプロンプトを選択的に保持することで一貫性と効率を高められるが、前提変更や利用回数を踏まえて保存範囲を設計する必要がある。"} Main layout: Two large rounded rectangular panels sit side by side beneath the lead sentence. The left panel is teal and titled “reasoning.context”; the right panel is olive green and titled “明示的プロンプトキャッシュ”. Connect the panels with a thin line and a small circular chain-link icon centered between them near the bottom, suggesting coordination between reasoning context and cache. A small upward green arrow near the right panel points back toward it. Left panel details: Include exactly 3 vertically stacked turn markers on the far left, connected by a vertical line with a downward arrow. The markers are labeled “1ターン”, “2ターン”, and “3ターン”. Next to each marker, show a small person icon and speech bubble. To the right, include exactly 3 horizontal rounded cards with teal icon blocks and Japanese explanatory text: 1) gear icon, “auto｜モデルの既定動作を利用”; 2) checklist icon, “all_turns｜安定した目標、前提、優先順位を複数ターンで再利用”; 3) clock icon, “current_turn｜以前の推論が不要になった場面で現在ターンに限定”. Right panel details: Include exactly 3 horizontal rounded cards with olive icon blocks and Japanese explanatory text: 1) document icon, “再利用する共通プレフィックスを開発者が指定”; 2) database cylinder icon, “キャッシュ書き込みは通常入力単価の1.25倍”; 3) percent tag icon, “読み出しは割引されるため、反復利用の回数が効果を左右する”. Bottom conclusion band: Across the bottom, create one wide rounded rectangle callout with a thin yellow border. On the left of the band place a gold target icon with an arrow. To the right, place a bold Japanese conclusion: {argument name="conclusion text" default="長期タスクの一貫性と効率を高める一方、前提変更時には過去の推論や不要なキャッシュを切り離す設計が必要"}. Visual style: Japanese corporate slide design, crisp vector line icons, rounded rectangles, light cream background, teal left section, olive right section, yellow conclusion accent, strong hierarchy, highly legible Japanese sans-serif typography, no decorative clutter. Constraints: Preserve the exact count of elements: 2 main panels, 3 left turn markers, 3 left option cards, 3 right cache cards, 1 central chain-link connector, 1 bottom conclusion callout, and 1 target icon. Do not add extra panels, extra cards, watermarks, logos, or decorative characters. Keep all slide text inside the infographic sharp and readable.