Infographic / Edu Visual - Layered GPT Image 2 Workflow Infographic

A dark futuristic Chinese infographic explaining layered AI image generation and compositing through a sci-fi sneaker advertisement example.

Prompt

4,886 charactersOpen Astorie studio ↗
Goal: Create a polished 16:9 Chinese tech infographic slide explaining the correct workflow for making complex design images with {argument name="AI image model name" default="GPT Image 2"}: layered generation plus compositing, using a futuristic sneaker advertisement as the example.

Canvas: Wide 16:9 dark interface-style presentation, approximately 1152×768, black-to-deep-navy background, thin glowing panel borders, rounded cards, subtle grid and glassmorphism. Accent colors are orange, cyan-blue, mint green, purple, and white. Overall look: premium AI design tutorial, SaaS dashboard, high contrast, crisp typography.

Header: At top left, large white title “GPT Image 2”. Under it, a two-color Chinese headline: orange text “复杂设计图的正确打开方式:” followed by white text “分层生成 + 合成”. Below, a smaller gray subtitle: “把复杂设计拆成三层分别生成,最后合成,效果更可控、细节更精准。”

Main layout: Split the slide into three columns. Left column compares wrong vs recommended method. Middle column shows exactly 3 separate prompt-layer example cards stacked vertically with plus signs between them. Right column shows the final composited sneaker poster. Bottom row contains exactly 2 comparison result cards on the left and exactly 3 recommendation cards on the right.

Left column details: Create 2 large rounded cards.
1. Top card, red warning style with red X icon. Title: “一次性生成(常见翻车原因)”. Body: “把所有元素塞进一个 Prompt”. Smaller explanation: “结果:构图 OK,但细节容易混乱,元素可能缺失或变形。” Include a small preview image of a futuristic running shoe floating over a sci-fi landscape with large Chinese text “未来已来” and tiny English “FUTURE IS NOW”.
2. Bottom card, green success style with green check icon. Title: “分层生成(推荐做法)”. Body: “每层只聚焦一件事”. Smaller explanation: “结果:每层精准可控,合成后整体效果更专业。” Inside this card, show exactly 3 mini layer rows with icons and labels: blue image icon row “背景层:场景 / 光线 / 氛围” with note “只描述环境、光照、色调、氛围”; purple cube icon row “主体层:角色 / 产品 / 动作” with note “只描述主体、材质、细节、姿态”; orange T icon row “文字装饰层:文案 / 图形 / 装饰” with note “只描述文字、排版、装饰视觉元素”.

Middle column details: Show exactly 3 stacked prompt example cards connected by cyan plus signs, and a cyan arrow pointing from the stack to the final poster on the right.
1. Blue card numbered “1”, title “背景层 Prompt 示例”. Text: “未来感科技场景,冷色调,日出光线,地面有反光,远处有星球”. Image: a wide empty sci-fi landscape with reflective water, mountains, sunrise, a large glowing ring, and a planet, no shoe.
2. Purple card numbered “2”, title “主体层 Prompt 示例”. Text: “科技感运动鞋,悬浮,侧视角,高细节,产品摄影”. Image: a single gray futuristic sneaker isolated on a transparent/checkered or pale background, no scenery, no text.
3. Orange card numbered “3”, title “文字装饰层 Prompt 示例”. Text: “科技风标题排版,中文主标题,英文副标题,简洁装饰线条”. Image: typography strip with Chinese title “未来已来”, English subtitle “FUTURE IS NOW”, small text “轻盈 / 科技 / 突破”, and fine orange decorative lines.

Right column final poster: Large vertical rounded panel with a green header strip. Header title: “合成后最终效果”. Header subtitle: “三层合成,整体精准,细节丰富,视觉统一”. Below it, create a cinematic finished advertisement: a detailed gray futuristic running shoe floating above reflective shallow water in a sci-fi alien landscape, with mountains and rocks, dawn light, a huge glowing circular portal/ring behind the shoe, and a large moon/planet in the upper right. At the bottom of the poster, include bold Chinese headline {argument name="Chinese headline text" default="未来已来"}, English subtitle {argument name="English subtitle text" default="FUTURE IS NOW"}, and small spaced feature words “轻盈 / 科技 / 突破”.

Bottom row: Create exactly 5 small cards total.
Left section titled “实测对比” with caption “同样的需求,不同的生成方式,结果差异明显。” It contains 2 cards: red X card “一次出全部” with notes “构图 OK,但细节乱,文字可能出错,元素可能变形” and a tiny preview; green check card “分三层合成” with notes “每层精准可控,合成后整体效果更专业” and a tiny preview.
Right section titled “使用建议” contains exactly 3 recommendation cards with icons: blue lightning icon “先用 Instant 模式快速验证思路(成本低、速度快)”; purple brain icon “确认结构后切 Thinking 模式生成高质量最终图”; green stacked-layers icon “分层结构 + 合成 = {argument name="AI image model name repeated" default="GPT Image 2"} 的真正用法”.

Visual style: Modern Chinese infographic, dark glass panels, glowing neon outlines, realistic 3D product render previews, crisp white Chinese typography, orange emphasis for the main concept, green for recommended workflow, red for mistakes, cyan arrows and connectors. Keep all text legible and aligned. Use the futuristic sneaker theme consistently; the shoe should be silver-gray, aerodynamic, high-detail, with sculpted sole holes and visible laces.

Constraints: Use exactly 3 layer example cards, exactly 3 mini layer rows, exactly 2 method comparison cards, exactly 3 usage recommendation cards, exactly 1 large final poster, and exactly 1 cyan arrow from the layer stack to the final poster. Do not add extra logos, watermarks, QR codes, people, or unrelated objects.
From sourceView source ↗

More to explore

Game Asset - Warhammer Scene: Rylanor vs Fulgrim

Game Asset - Warhammer Scene: Rylanor vs Fulgrim

Mainstream MadnessGPT Image 2
Ancient Rylanor, the Ancient of Rites of the Emperor's Children, makes his final stand deep beneath the burned world of Istvaan III. He is facing Fulgrim, the Daemon Primarch of Slaanesh.
He stands defiant, battered and half-buried in the ruins of a subterranean vault, his surviving honourable soul burning with hatred for the Traitors. Around him lie the bones of his long-dead loyalist brothers.
Game Asset - Halloween Robot at Tokyo Gym

Game Asset - Halloween Robot at Tokyo Gym

カーブミラーGPT Image 2
A photorealistic, vertical full-body shot of a massive, bipedal humanoid robot standing outdoors in front of a modern Japanese municipal sports hall. The robot is constructed from heavy-duty industrial machinery, featuring segmented white and black armor plating, exposed hydraulic pistons, and complex mechanical joints. Its head is a rectangular sensor array with multiple camera lenses and glowing red emergency lights on the shoulders. The robot is decorated for Halloween: it wears a flowing orange cape patterned with black bats and jack-o'-lanterns, and its body is draped with garlands of purple tinsel, small plastic pumpkins, and triangular bunting flags. Several chest plates are labeled with the text "警視庁" (Metropolitan Police Department) and "MPD".

In the foreground, two police officers wearing dark blue uniforms, tactical vests, and helmets stand with their backs to the viewer; their vests also bear the text "警視庁 MPD". To the right, a crowd of civilians includes a woman taking a photo with her smartphone and a child dressed in an orange witch costume with a pointed hat. The background features a large concrete building with glass windows and prominent signage reading "区立総合体育館" (Ward General Gymnasium). A tall vertical banner hangs on the right side of the building, displaying a cartoon pumpkin and the text "ハロウィン スポーツ フェスティバル" (Halloween Sports Festival). The scene is lit by bright, natural daylight under a clear blue sky.
Product Marketing - Olive Green Waxed Canvas Roll-Top Backpack

Product Marketing - Olive Green Waxed Canvas Roll-Top Backpack

SjoltGPT Image 2
A high-resolution product photograph of a dark olive green waxed canvas roll-top backpack. The bag features two vertical tan leather straps with square brass buckles on the front, and a horizontal zippered pocket below them with a brass zipper pull. A matching fabric grab handle is visible at the top center. The backpack sits upright on a light beige stone surface against a plain, textured cream-colored wall background. Soft, diffused daylight illuminates the scene from the left, casting gentle shadows to the right.
Profile / Avatar - Retro American Cartoon Style Image Transformation

Profile / Avatar - Retro American Cartoon Style Image Transformation

SairahGPT Image 2
Transform the exact uploaded image into a full body illustration designed in a 1960s–1980s retro American cartoon style. Preserve the exact subject, facial identity, hairstyle, clothing, accessories, colors, proportions, pose, body language, facial expression, camera angle, framing, composition, perspective, lighting, and every visible detail from the original image. Do not redesign, reinterpret, simplify, replace, or invent any elements. The uploaded image must remain instantly recognizable, changing only the artistic style.

The character should feature exaggerated rubber hose proportions, expressive body language, a slightly oversized head, simplified anatomy, elastic limbs, large classic oval cartoon eyes with simple black pupils, thin expressive eyebrows, rounded hands, clean geometric shapes, and a playful vintage personality while preserving the exact likeness, expression, and silhouette of the original subject.

The illustration features thick bold black outlines, flat solid colors, minimal cel-style shading, subtle vintage film grain texture, slightly desaturated tones like old animation cels, smooth hand-drawn 2D linework, simple color blocking, soft retro print texture, and a classic mid-century American animation aesthetic.

Preserve the original background exactly as shown in the uploaded image unless otherwise specified.

Keep the original composition, framing, perspective, and camera angle exactly the same. Full body should remain visible if it is visible in the original image. High resolution, clean illustration, no realism, no 3D rendering, no anime, no Pixar, no Disney, no modern cartoon style, only authentic 1960s–1980s American retro animation aesthetics.
Profile / Avatar - Retro Showa Halloween Portrait

Profile / Avatar - Retro Showa Halloween Portrait

いがぐりメタボGPT Image 2
A vertical, medium-shot portrait of a young Asian woman with short dark hair and bangs, wearing a homemade Halloween costume consisting of an orange paper pumpkin hat and a matching jagged orange paper collar tied at the neck over a black long-sleeved dress. She is leaning slightly forward, making a peace sign with her right hand near her cheek, while holding a small brown paper bag decorated with a jack-o'-lantern and black cat in her left hand. The background is a narrow, dimly lit retro Japanese alleyway (yokocho) at night, featuring wooden storefronts, glowing lanterns, and blurred figures of other people walking away. On the left, there are hanging paper decorations including a jack-o'-lantern and a white ghost. On the right, a vertical banner reads 'おもちや' (Omo-cha). The image has a strong Showa-era retro aesthetic, characterized by warm, faded colors, visible film grain, scratches, dust spots, and soft focus typical of vintage instant photography.
Profile / Avatar - Retro Halloween Vampire Girl

Profile / Avatar - Retro Halloween Vampire Girl

いがぐりメタボGPT Image 2
A retro-style photograph of a young Asian woman with short, messy dark hair sitting in a vintage floral armchair. She is wearing a beige chunky knit sweater over a plaid skirt, paired with slouchy white socks and brown loafers. A black velvet vampire cape with red satin lining is draped over her shoulders, tied at the neck with a ribbon. Her hands are raised in a playful claw gesture, and she has small fangs visible. The background features Halloween decorations including orange jack-o'-lanterns and black bat cutouts on the wall, along with an old CRT television set to the left. The image has a nostalgic Showa-era aesthetic with warm lighting and film grain.

Find your next direction

ASTORIE CANVAS

Take the idea further.

Bring your prompts, references and results together in one creative space.

Open your canvas ↗