耗时100小时掌握ChatGPT图像生成器，这是我学到的一切

result = client.images.edit(
    model="gpt-image-1.5",
    input_fidelity="high",  # 身份保留的秘诀
    quality="high",
    image=[open("portrait.png", "rb")],
    prompt="Change the background to a sunset beach while preserving the person's exact appearance"
)

这种组合确保在应用你请求的更改时最大程度地保留原始主体。

Goal/Output (目标/输出):
- [Type of image: ad, UI mockup, infographic, photo, illustration] (图像类型：广告、UI模型、信息图、照片、插图)
- [Intended use and audience] (预期用途和受众)

Scene (场景):
- [Background/environment description] (背景/环境描述)
- [Main subject with specific details] (主体及具体细节)
- [Action or relationship between elements] (动作或元素间的关系)

Style (风格):
- [Medium: photograph, watercolor, 3D render, vector illustration] (媒介：照片、水彩、3D渲染、矢量插图)
- [Key textures: matte, glossy, grainy, smooth, organic] (关键纹理：哑光、光泽、颗粒感、平滑、有机)
- [Quality descriptors: realistic imperfections, stylized, minimalist] (质量描述：逼真的瑕疵、风格化、极简主义)

Composition/Layout (构图/布局):
- [Camera position: close-up, wide shot, aerial view, eye-level] (相机位置：特写、广角、鸟瞰、平视)
- [Lighting: golden hour, studio strobes, overcast, dramatic shadows] (光照：黄金时刻、摄影棚闪光灯、阴天、戏剧性阴影)
- [Element placement: centered, rule of thirds, negative space, margins] (元素放置：居中、三分法、留白、边距)

Text (if any) (文本，如果有):
- "Exact text in quotes" ("引号中的确切文本")
- [Font style, size, color, position] (字体风格、大小、颜色、位置)
- [Specify: render only once, no duplicates] (指定：仅渲染一次，无重复)

Constraints (约束):
- Change ONLY: [specific element if editing] (仅更改：[特定元素，如果是编辑])
- Preserve exactly: [elements that must stay unchanged] (完全保留：[必须保持不变的元素])
- Negative: no watermark, no extra text, no logos, no [unwanted elements] (负向：无水印、无额外文本、无Logo、无[不需要的元素])

这个框架为模型提供了它需要做出的每一个视觉决策的清晰上下文。

❌ BEFORE (模糊):
"A beautiful portrait of an old fisherman, very detailed, high quality, realistic"
(一个老渔夫的美丽肖像，非常详细，高质量，逼真)

✅ AFTER (摄影师思维):
"Candid documentary photograph of an elderly fisherman on a weathered wooden boat.
Weathered face with visible wrinkles, sun spots, and pores. Deep-set kind eyes.
Gray stubble. Faded traditional anchor tattoo on forearm. Salt-stained navy wool
sweater, worn cap.

Early morning coastal light, soft fog diffusing the sun. Medium close-up at eye
level, 50mm lens, f/2.8, shallow depth of field. Shot like 35mm film with subtle
grain, natural color balance.

Documentary style — honest, unretouched, capturing a real moment. No glamorization."
(饱经风霜的木船上的老渔夫的纪实抓拍。饱经风霜的脸，可见皱纹、晒斑和毛孔。深陷的慈祥眼睛。灰色胡茬。前臂上褪色的传统锚纹身。盐渍的海军蓝羊毛衫，磨损的帽子。
清晨的海岸光线，柔和的雾气漫射着阳光。视线水平的中特写，50mm镜头，f/2.8，浅景深。像35mm胶片一样拍摄，有微妙的颗粒感，自然色彩平衡。
纪实风格 —— 诚实，未修饰，捕捉真实瞬间。没有美化。)

摄影师思维将模糊的愿望转化为模型能够深度解码的精确视觉规格。

Create a photorealistic candid photograph of an elderly sailor standing on a small fishing boat.
(创建一张老水手站在小渔船上的照片级真实抓拍。)

Subject: Weathered face with visible wrinkles, sun spots, and pores. Deep-set kind
eyes with crow's feet. Gray stubble, a few days unshaven. Faded traditional anchor
tattoo on forearm. Salt-stained navy wool sweater, worn and pilled. Creased cap
with faded insignia.
(主体：饱经风霜的脸，可见皱纹、晒斑和毛孔。深陷的慈祥眼睛，有鱼尾纹。灰色胡茬，几天没刮。前臂上褪色的传统锚纹身。盐渍的海军蓝羊毛衫，磨损起球。折痕帽子，褪色徽章。)

Setting: Early morning on the water, soft coastal fog diffusing the light. Aged
wooden boat deck with peeling paint, fishing nets in background, coiled rope.
(环境：清晨的水面，柔和的海岸雾气漫射着光线。陈旧的木船甲板，油漆剥落，背景中有渔网，盘绕的绳索。)

Technical: Shot like 35mm film photography, medium close-up at eye level, 50mm
lens, shallow depth of field with boat blurred behind him. Subtle film grain,
natural color balance without heavy grading.
(技术：像35mm胶片摄影一样拍摄，视线水平的中特写，50mm镜头，浅景深，身后船只模糊。微妙的胶片颗粒，自然色彩平衡，无重度调色。)

The image should feel like a real moment captured by a photojournalist — honest,
unposed, with real skin texture, worn materials, and everyday imperfection. No
glamorization, no heavy retouching, no artificial perfection.
(图像应该感觉像是摄影记者捕捉到的真实瞬间 —— 诚实、未摆拍，有真实的皮肤纹理、磨损的材料和日常的瑕疵。没有美化，没有重度修饰，没有人工完美。)

注意我们要如何明确要求瑕疵 —— 饱经风霜的皮肤、磨损的材料、剥落的油漆。现实是有纹理的。

Create a detailed infographic explaining how a coffee machine works.
(创建一个详细的信息图，解释咖啡机是如何工作的。)

Structure (结构):
- Title at top: "The Journey of Your Morning Coffee" (顶部标题："你晨间咖啡的旅程")
- Vertical flow diagram showing: bean hopper → grinder → portafilter →
  grouphead → water heating → extraction → cup
  (垂直流程图显示：豆仓 → 研磨机 → 手柄 → 冲煮头 → 水加热 → 萃取 → 杯子)
- Each step has an icon and 1-2 sentence explanation (每一步都有一个图标和1-2句话的解释)
- Warm color palette (browns, creams, copper accents) (暖色调：棕色、奶油色、铜色点缀)
- Clean, modern design with plenty of white space (干净、现代的设计，大量留白)
- Subtle coffee stain texture in background corners (背景角落有微妙的咖啡渍纹理)

Style (风格): Professional print-quality infographic, vector-style icons, clear
hierarchy, readable at A4 size. (专业印刷质量信息图，矢量风格图标，清晰层级，A4尺寸可读。)

Typography (排版): Clean sans-serif headings, readable body text, clear visual
hierarchy between title, section headers, and explanatory text. (干净的无衬线标题，可读的正文，标题、章节头和解释性文本之间有清晰的视觉层级。)

No watermarks. No stock photo elements. Original illustration only. (无水印。无图库照片元素。仅原创插图。)

对于密集的文本和复杂的布局，始终使用 quality="high" 以确保文本保持清晰可读。

Create an original logo for "Field & Flour" — a local artisan bakery.
(为"Field & Flour" —— 一家当地手工面包店创建一个原创Logo。)

Brand personality: Warm, authentic, handcrafted, timeless. Not trendy or corporate.
(品牌个性：温暖、真实、手工制作、永恒。不追逐潮流或企业化。)

Design requirements (设计要求):
- Clean vector-style shapes with strong silhouette (干净的矢量风格形状，轮廓强)
- Balanced negative space (平衡的负空间)
- Must read clearly from 16px favicon to large signage (从16px网站图标到大型标牌都必须清晰可读)
- Flat design, minimal strokes, no gradients unless essential (扁平化设计，极简线条，除非必要否则无渐变)
- Earth-tone palette: warm wheat gold, deep brown, cream (大地色调：暖麦金、深棕、奶油色)
- Could incorporate subtle wheat or grain element (可以融入微妙的小麦或谷物元素)
- Text must be perfectly legible and properly kerned (文本必须完全清晰且字距适当)

Output: Single centered logo on plain cream background. Generous padding around
the design for flexibility. (输出：纯奶油色背景上的单个居中Logo。设计周围有宽敞的留白以增加灵活性。)

No watermarks, no mockups, no 3D effects, no complex imagery. Simple, functional,
timeless design. (无水印，无样机，无3D效果，无复杂图像。简单、实用、永恒的设计。)

使用 n=4 生成多个变体。Logo设计是主观的 —— 给自己多种选择。

Create a realistic mobile app UI mockup for a local farmers market app.
(为一个当地农贸市场App创建一个逼真的移动App UI模型。)

Screen content (from top) (屏幕内容（从上到下）):
- Simple header with market name "Riverside Market" and search icon (简单的头部，带有市场名称"Riverside Market"和搜索图标)
- Today's featured vendor carousel with square photos (今日精选商家轮播，带方形照片)
- "Fresh Today" section with produce category chips (Vegetables, Fruits, Dairy, Baked) ("今日新鲜"部分，带农产品类别标签（蔬菜、水果、乳制品、烘焙）)
- Vendor list with small photos, names, specialties, and distance (商家列表，带小照片、名称、特色和距离)
- Bottom navigation: Home, Map, Favorites, Cart, Profile (底部导航：首页、地图、收藏、购物车、个人资料)

Design language (设计语言):
- White background, subtle natural green accents (白色背景，微妙的自然绿色点缀)
- Clear typography hierarchy (system fonts feel) (清晰的排版层级（系统字体感觉）)
- Generous padding and touch-friendly targets (宽敞的填充和触摸友好的目标)
- Looks like a real shipped product, not a concept (看起来像真正的已发布产品，而不是概念)
- Uses realistic vendor names and produce photos (使用真实的商家名称和农产品照片)

Frame: Place the UI inside an iPhone 15 Pro device frame, slight perspective
tilt, subtle shadow beneath. (框架：将UI放置在iPhone 15 Pro设备框架内，轻微透视倾斜，下方有微妙阴影。)

专注于布局、层级、间距和逼真的界面元素。避免概念性或艺术性的语言。

Create a 4-panel vertical comic strip. Equal panel sizes, clear panel borders.
(创建一个4格垂直连环画。格子大小相等，清晰的边框。)

Panel 1: Pet owner walks out the front door, keys in hand. Through the window
behind them, we see their cat watching — paws pressed against glass, eyes wide
with apparent sadness. The house suddenly feels empty.
(第一格：宠物主人走出前门，手里拿着钥匙。透过身后的窗户，我们看到他们的猫在看 —— 爪子按在玻璃上，眼睛睁得大大的，明显很悲伤。房子突然感觉空荡荡的。)

Panel 2: The door clicks shut. The cat slowly turns away from the window toward
the empty house. Its posture shifts from forlorn to interested. Eyes narrow with
possibility.
(第二格：门关上了。猫慢慢从窗户转向空荡荡的房子。它的姿态从凄凉转变为感兴趣。眼睛眯起来，充满了可能性。)

Panel 3: Total chaos. Cat sprawled across the forbidden couch like royalty.
Knocked over plant on the floor. Papers scattered. Sunbeam spotlighting the
scene of domestic crime.
(第三格：完全混乱。猫像皇室一样躺在禁止入内的沙发上。地板上是被打翻的植物。纸张散落。阳光照亮了家庭犯罪现场。)

Panel 4: Door handle turns. Cat sits perfectly upright by the entrance,
composed and innocent, tail wrapped neatly around paws. Not a hair out of
place. As if nothing happened.
(第四格：门把手转动。猫在入口处坐得笔直，镇定而无辜，尾巴整齐地缠绕在爪子上。一丝不乱。仿佛什么都没发生过。)

Style: Warm illustrated style with expressive characters, clear visual
storytelling that reads without text. Consistent character design across
all panels.
(风格：温暖的插图风格，富有表现力的角色，清晰的视觉叙事，无需文字即可阅读。所有格子中角色设计一致。)

No speech bubbles or text. Let the visuals tell the story. (无气泡或文字。让视觉讲述故事。)

将每一格定义为具有清晰动作的独特视觉节拍。模型处理格子布局和视觉连续性。

Create a children's book illustration introducing the main character.
(创建一个介绍主角的儿童读物插图。)

Character: Young forest hero, around 8 years old. (角色：年轻的森林英雄，约8岁。)
- Green hooded tunic (think woodland adventurer, not Robin Hood) (绿色连帽长袍（想想林地冒险家，不是罗宾汉）)
- Soft brown boots, well-worn (柔软的棕色靴子，磨损良好)
- Small belt pouch for collecting treasures (用于收集宝藏的小腰包)
- Carries a tiny wooden bow (symbolic, for helping not hurting) (背着一把小木弓（象征性的，用于帮助而非伤害）)
- Kind expression, bright curious eyes, brave but gentle demeanor (善良的表情，明亮好奇的眼睛，勇敢但温和的举止)
- Slightly oversized head for picture book proportions (略大的头，符合图画书比例)

Theme: This character protects and rescues small forest animals in trouble.
(主题：这个角色保护和拯救陷入困境的小森林动物。)

Style: Hand-painted watercolor look with soft outlines, warm earthy palette
with forest greens and autumn oranges. Whimsical, friendly, inviting for
young readers ages 4-8.
(风格：手绘水彩外观，柔和轮廓，温暖的大地色调，森林绿和秋季橙。异想天开，友好，吸引4-8岁的年轻读者。)

Composition: Character standing in simple forest glade, dappled sunlight,
leaving room for title text above. Character clearly showcased.
(构图：角色站在简单的林间空地上，斑驳的阳光，上方留有标题文字空间。角色清晰展示。)

Original character design only. No text. No watermarks. No copyrighted
character references. (仅原创角色设计。无文字。无水印。无版权角色参考。)

保存此角色参考图像 —— 你将使用它在后续插图中保持一致性。

Create a realistic outdoor crowd scene in Bethel, New York on August 16, 1969.
(创建一个1969年8月16日纽约伯特利的逼真户外人群场景。)

Photorealistic, period-accurate clothing, staging, and environment.
(照片级真实，符合时代的服装、布景和环境。)

Documentary photography style, shot on film, natural lighting.
(纪实摄影风格，胶片拍摄，自然光。)

模型知道这是伍德斯托克音乐节，而无需被告知。它仅根据日期和地点就能生成嬉皮士、那个时代的时尚、节日气氛。

Edit the image to dress this person in the provided clothing items.
(编辑图像，让此人穿上提供的服装项目。)

MUST PRESERVE (do not change in any way) (必须保留（不要以任何方式更改）):
- Face, facial features, expression, skin tone (脸、面部特征、表情、肤色)
- Body shape, proportions, and pose (体型、比例和姿势)
- Hairstyle and hair color (发型和发色)
- Background and environment (背景和环境)
- Camera angle, framing, and composition (相机角度、取景和构图)
- Overall lighting direction and quality (整体光照方向和质量)

CHANGE ONLY (仅更改):
- Replace current clothing with provided garment images (用提供的服装图像替换当前服装)
- Fit garments naturally to body geometry (使服装自然贴合身体几何形状)
- Show realistic fabric draping, folds, and behavior (显示逼真的织物垂坠、褶皱和表现)
- Match lighting and shadows on fabric to original photo (使织物上的光照和阴影与原始照片匹配)

REQUIREMENTS (要求):
- Photorealistic integration — outfit should look worn, not pasted (照片级真实融合 —— 服装应该看起来是穿在身上的，而不是粘贴上去的)
- Maintain color temperature of original image (保持原始图像的色温)
- No accessories, text, logos, or watermarks added (不添加配饰、文字、Logo或水印)
- Identity must remain clearly recognizable (身份必须保持清晰可辨)

对于虚拟试穿，始终使用 input_fidelity="high" 以确保面部相似度得到保持。

Using the EXACT visual style of the reference image (Image 1), create:
(使用参考图像（图像1）的完全相同的视觉风格，创建：)
A man riding a motorcycle on a winding mountain road.
(一个男人在蜿蜒的山路上骑摩托车。)

STYLE ELEMENTS TO MATCH PRECISELY from reference (从参考中精确匹配的风格元素):
- Color palette and saturation levels (调色板和饱和度水平)
- Line quality and weight (线条质量和粗细)
- Texture treatment and brushwork (纹理处理和笔触)
- Lighting style and direction (光照风格和方向)
- Level of detail vs. abstraction (细节与抽象的程度)
- Overall artistic aesthetic (整体艺术审美)

APPLY TO NEW CONTENT (应用到新内容):
- Single subject (man on motorcycle) (单一主体（骑摩托车的人）)
- Clear composition with visual interest (清晰的构图，具有视觉趣味)
- Mountain road environment with curves (带有弯道的山路环境)
- Sense of motion and freedom (运动感和自由感)

The new image should look like it came from the same artist or series as
the reference. Maintain stylistic consistency exactly.
(新图像应该看起来像与参考图像出自同一位艺术家或系列。精确保持风格一致性。)

当你具体说明要保留哪些风格元素以及要更改哪些内容元素时，风格迁移效果最好。

In this room photo, replace ONLY the white plastic chairs with
mid-century modern wooden chairs (walnut finish, tapered legs,
woven seat).
(在这个房间照片中，仅将白色塑料椅子替换为世纪中期现代风格的木椅（胡桃木饰面，锥形腿，编织座椅）。)

PRESERVE COMPLETELY (完全保留):
- Camera angle and perspective (相机角度和透视)
- Room lighting direction and quality (房间光照方向和质量)
- All other furniture and objects (所有其他家具和物体)
- Wall colors and decorations (墙壁颜色和装饰)
- Floor material and shadows (地板材质和阴影)
- Overall image quality and color grading (整体图像质量和调色)

INTEGRATION REQUIREMENTS (融合要求):
- Chairs must match room's perspective exactly (椅子必须精确匹配房间的透视)
- Wood grain should catch existing light realistically (木纹应逼真地捕捉现有光线)
- Contact shadows must be natural and match light source (接触阴影必须自然并与光源匹配)
- Scale must be accurate relative to table height (比例必须相对于桌子高度准确)
- New chairs should look like they belong in this room (新椅子应该看起来属于这个房间)

Photorealistic result — should look like the original photograph.
(照片级真实结果 —— 应该看起来像原始照片。)

室内设计可视化是最具商业价值的编辑应用之一。

Transform this hand-drawn sketch into a photorealistic image.
(将这张手绘草图转化为照片级真实的图像。)

PRESERVE FROM SKETCH (从草图中保留):
- Exact layout and proportions (精确的布局和比例)
- Perspective and viewing angle (透视和视角)
- Element placement and relationships (元素放置和关系)
- Implied depth and layering (暗示的深度和层次)

ADD FOR REALISM (为真实感添加):
- Appropriate real-world materials and textures (适当的现实世界材料和纹理)
- Consistent natural lighting (interpret from sketch shading) (一致的自然光照（从草图阴影解读）)
- Environmental context matching the implied setting (匹配暗示环境的环境背景)
- Surface imperfections and wear appropriate to materials (适合材料的表面瑕疵和磨损)

CONSTRAINTS (约束):
- Do not add new elements not present in sketch (不要添加草图中不存在的新元素)
- Do not add text or watermarks (不要添加文字或水印)
- Treat the sketch as an architectural blueprint to follow exactly (将草图视为要精确遵循的建筑蓝图)
- Fill in realistic details while honoring the original composition (在尊重原始构图的同时填充逼真的细节)

模型解读草图的意图并填充逼真的细节，同时保持原始构图。

Transform this daytime summer scene into a winter evening with snowfall.
(将这个白天的夏季场景转换为下雪的冬夜。)

CHANGE (更改):
- Time of day: from afternoon to dusk (warm interior lights visible) (时间：从下午到黄昏（可见温暖的室内灯光）)
- Season: summer to deep winter (季节：夏季到深冬)
- Weather: clear to active snowfall (天气：晴朗到下雪)
- Ground: grass to fresh snow coverage (地面：草地到新雪覆盖)
- Trees: summer foliage to bare branches with snow (树木：夏季树叶到带雪的枯枝)
- Atmosphere: add visible breath if people present (氛围：如果有人，添加可见的呼吸)
- Surfaces: add frost on windows and metal (表面：在窗户和金属上添加霜)

PRESERVE (保留):
- Camera position and angle exactly (相机位置和角度完全不变)
- All objects and their exact positions (所有物体及其确切位置)
- Architecture and structural elements (建筑和结构元素)
- People and their poses (update clothing appropriately) (人和他们的姿势（适当更新服装）)
- Overall composition and framing (整体构图和取景)

Style: Photorealistic, natural atmospheric perspective, visible
snowflakes in air, cozy contrast between warm interior lights and
cold exterior. Should feel photographed, not filtered.
(风格：照片级真实，自然大气透视，空气中可见雪花，温暖室内灯光与寒冷室外之间的舒适对比。应该感觉是拍摄的，而不是滤镜处理的。)

在环境转换中使用 input_fidelity="high" 和 quality="high" 以获得最佳结果。

I'm providing 2 images:
(我提供2张图片：)
- Image 1: Beach scene with woman standing on shore at sunset (图片1：日落时分女人站在岸边的海滩场景)
- Image 2: Golden retriever sitting in a studio setting (图片2：坐在摄影棚环境中的金毛寻回犬)

Task: Place the dog from Image 2 into the beach scene from Image 1,
positioned next to the woman, looking up at her.
(任务：将图片2中的狗放入图片1的海滩场景中，位于女人旁边，抬头看着她。)

MATCHING REQUIREMENTS (匹配要求):
- Dog's lighting must match beach sunset (warm golden light from left) (狗的光照必须匹配海滩日落（来自左侧的温暖金光）)
- Scale dog appropriately relative to woman's height (相对于女人的身高适当缩放狗)
- Dog should cast shadow consistent with scene's sun angle (狗应该投射与场景太阳角度一致的阴影)
- Sand texture should show around and under dog's paws (狗爪周围和下方应显示沙子纹理)
- Fur should catch the same golden hour highlights as scene (毛发应捕捉与场景相同的黄金时刻高光)

PRESERVE FROM IMAGE 1 (从图片1保留):
- Woman's exact appearance, position, and pose (女人的确切外貌、位置和姿势)
- Beach background completely unchanged (海滩背景完全不变)
- Original photo's color grading and mood (原始照片的调色和情绪)

The composite should look like a single photograph taken on location.
No visible compositing artifacts.
(合成图应该看起来像是在现场拍摄的单张照片。没有可见的合成伪影。)

按编号引用图片，并明确哪些元素转移，哪些保持固定。

Translate all text in this infographic from English to Japanese.
(将此信息图中的所有文本从英语翻译成日语。)

MUST PRESERVE (必须保留):
- Exact layout, spacing, and positioning of all elements (所有元素的精确布局、间距和位置)
- All visual elements, icons, illustrations, and graphics (所有视觉元素、图标、插图和图形)
- Typography hierarchy (headlines vs body text relationships) (排版层级（标题与正文关系）)
- Color scheme and overall design aesthetic (配色方案和整体设计审美)
- Font weights and relative sizes (字体粗细和相对大小)

TRANSLATION REQUIREMENTS (翻译要求):
- Accurate Japanese translation with natural phrasing (准确的日语翻译，措辞自然)
- Match visual weight and style to original fonts (匹配原始字体的视觉权重和风格)
- Adjust character spacing for Japanese typographic norms (针对日语排版规范调整字符间距)
- No text truncation or overflow outside original bounds (无文本截断或溢出原始边界)

Do not modify any non-text elements. Only change the language.
(不要修改任何非文本元素。只改变语言。)

此工作流无需从头重建即可处理营销材料、UI截图、包装和信息图表。

Continue the children's book story using the character from the reference image.
(使用参考图像中的角色继续儿童读物故事。)

New Scene (新场景):
The same young forest hero is gently helping a frightened squirrel out
of a fallen hollow tree after a winter storm. Snow on the ground, bare
branches above, warm light filtering through clouds.
(同一个年轻的森林英雄在一场冬季风暴后，正温柔地帮助一只受惊的松鼠从倒下的空心树中出来。地面上有雪，头顶是光秃秃的树枝，温暖的光线透过云层过滤下来。)

CHARACTER CONSISTENCY (from reference) (角色一致性（来自参考）):
- Same green hooded tunic, exact shade and style (同样的绿色连帽长袍，确切的色调和款式)
- Same soft brown boots (同样的柔软棕色靴子)
- Same belt pouch (同样的腰包)
- Same facial features, proportions, and color palette (同样的面部特征、比例和调色板)
- Same gentle, heroic personality in expression (表情中同样的温和、英雄个性)
- Same children's book proportions (同样的儿童读物比例)

STYLE CONSISTENCY (from reference) (风格一致性（来自参考）):
- Same watercolor illustration style (同样的水彩插图风格)
- Same soft outlines (同样的柔和轮廓)
- Same warm earthy color treatment (同样的温暖大地色处理)
- Same whimsical, friendly aesthetic (同样的异想天开、友好审美)

New elements: winter forest environment, frightened squirrel, fallen
tree with hollow. (新元素：冬季森林环境，受惊的松鼠，倒下的空心树。)

Do not redesign the character. Do not change the artistic style.
No text. No watermarks.
(不要重新设计角色。不要改变艺术风格。无文字。无水印。)

引用锚点图像并重复关键角色细节，以在整本书中保持一致性。

Create a hyper-stylized 3D floating head portrait based on this person.
(根据此人创建一个超风格化的3D浮动头像人像。)

STYLE CHARACTERISTICS (风格特征):
- Smooth skin with glossy vinyl-finish surface (光滑皮肤，带有光泽乙烯基表面)
- Strong highlighter on cheekbones and nose tip catching soft light (颧骨和鼻尖上的强高光捕捉柔和光线)
- Holographic, iridescent eyeshadow (purple to teal color shift) (全息、彩虹色眼影（紫色到青色变色）)
- Thick hair sculpted in slick, glossy waves like polished acrylic (像抛光亚克力一样雕刻成光滑、光泽波浪的浓密头发)
- Small metallic chrome nose piercing with brushed reflections (带有拉丝反射的小金属铬鼻钉)

EXPRESSION (表情):
Confident, slightly unimpressed look — half-lidded eyes, subtly
arched brow, the sophisticated "too cool" attitude.
(自信，略带不屑的表情 —— 半垂的眼睑，微妙拱起的眉毛，复杂的“太酷了”的态度。)

TECHNICAL SPECIFICATIONS (技术规格):
- Head floats isolated against plain white background (头像孤立漂浮在纯白背景前)
- Slight 15-degree tilt (premium product render feeling) (轻微15度倾斜（高级产品渲染感）)
- Bright, diffuse studio lighting with no harsh shadows (明亮、漫射的摄影棚灯光，无刺眼阴影)
- Emphasis on glossy, plastic, subsurface scattering effects (强调光泽、塑料、次表面散射效果)
- Ultra-smooth textures throughout (整体超平滑纹理)
- Close-up portrait angle, straight-on, 85mm lens feel (特写人像角度，正视，85mm镜头感)

The result should look like a high-end 3D character render or
collectible figure — plastic perfection with personality.
(结果应该看起来像高端3D角色渲染或收藏人偶 —— 具有个性的塑料完美感。)

这种程度的审美细节能在不同主体间产生惊人一致的结果。

Transform this person into an adorable chibi-style character.
(将此人转换为可爱的Q版风格角色。)

CHIBI PROPORTIONS (Q版比例):
- Tiny body (about 1 head-height tall) (微小的身体（约1个头高）)
- Oversized head (3x body proportions) (超大的头（3倍身体比例）)
- Large, sparkling eyes with cute highlights (大而闪亮的眼睛，带有可爱的反光)
- Soft, rounded facial features (柔和、圆润的面部特征)
- Cheerful, expressive pose with personality (愉快、富有表现力的姿势，具有个性)

PRESERVE FROM ORIGINAL (从原始图像保留):
- Recognizable facial features (simplified but identifiable) (可辨认的面部特征（简化但可识别）)
- Hairstyle, length, and hair color (发型、长度和发色)
- Distinctive clothing style or accessories (独特的服装风格或配饰)
- Any notable characteristics (glasses, jewelry, etc.) (任何显著特征（眼镜、珠宝等）)
- Overall personality and vibe (整体个性和氛围)

STYLE (风格):
- Smooth pastel shading (平滑的柔和阴影)
- Clean lines and simplified details (干净的线条和简化的细节)
- Bright, expressive colors (明亮、富有表现力的颜色)
- Collectible figure aesthetic (收藏人偶审美)

Background: Simple gradient or plain color to showcase character.
(背景：简单的渐变或纯色以展示角色。)

The result should feel like an irresistible chibi mascot that
clearly represents the original person.
(结果应该感觉像是一个无法抗拒的Q版吉祥物，清楚地代表原始人物。)

Q版转换非常适合个人品牌、团队头像和商品设计。

Create a realistic highway billboard mockup featuring this product.
(创建一个以该产品为特色的逼真高速公路广告牌模型。)

BILLBOARD CONTENT (广告牌内容):
- Product bottle prominently displayed on left third (产品瓶子醒目地展示在左三分之一处)
- Main headline on right (EXACT TEXT, render verbatim): (右侧主标题（确切文本，逐字渲染）：)
  "Fresh & Clean — Every Day"
- Tagline below headline: "Nature's Best Ingredients" (标题下方的标语："Nature's Best Ingredients")
- Small logo placeholder area in bottom right corner (右下角的小Logo占位区域)

TYPOGRAPHY SPECIFICATIONS (排版规格):
- Headline: Bold sans-serif, white text, high contrast (标题：粗体无衬线，白色文本，高对比度)
- Tagline: Light sans-serif, slightly smaller, same white (标语：细无衬线，稍小，同样的白色)
- Clean kerning, centered alignment within text area (干净的字距，文本区域内居中对齐)
- Text appears EXACTLY ONCE — no duplicates anywhere (文本只出现一次 —— 任何地方都没有重复)

SCENE (场景):
- Billboard on highway overpass or roadside structure (高速公路立交桥或路边结构上的广告牌)
- Sunset lighting creating warm, appealing atmosphere (日落光线营造温暖、吸引人的氛围)
- Photorealistic environment with motion-blurred vehicles below (照片级真实环境，下方有动态模糊的车辆)
- Professional advertising photography feel (专业广告摄影感)

No watermarks. No additional marketing copy. No logos unless
specified. Text must be perfectly legible and correctly spelled.
(无水印。无额外营销文案。除非指定，否则无Logo。文本必须完全清晰且拼写正确。)

对于包含文本的营销材料，始终使用 quality="high"。在最终使用前验证拼写。

Extract the product from this image for e-commerce use.
(从这张图像中提取产品用于电子商务。)

OUTPUT SPECIFICATIONS (输出规格):
- Transparent background (RGBA PNG format) (透明背景（RGBA PNG格式）)
- Crisp silhouette with clean edges (清晰的轮廓，边缘干净)
- No halos or color fringing around product (产品周围无光晕或彩色条纹)
- All product labels and text perfectly preserved (所有产品标签和文本完美保留)
- Exact product geometry and proportions maintained (精确保持产品几何形状和比例)

OPTIONAL ENHANCEMENT (可选增强):
- Add subtle, realistic contact shadow (添加微妙、逼真的接触阴影)
- Shadow should be soft and natural, no hard edges (阴影应柔和自然，无硬边)
- Shadow works with the transparent background (阴影适用于透明背景)

CRITICAL CONSTRAINTS (关键约束):
- Do NOT restyle or recolor the product (不要重新设计或重新着色产品)
- Do NOT modify product appearance in any way (不要以任何方式修改产品外观)
- Only remove background and add optional shadow (仅移除背景并添加可选阴影)
- Preserve every detail of the original product exactly (精确保留原始产品的每一个细节)

注意：当前模型渲染棋盘格图案来表示透明度 —— 可能需要后期处理才能获得真正的Alpha通道。

import os
import base64
from openai import OpenAI

client = OpenAI()

# Create output directory (创建输出目录)
os.makedirs("output_images", exist_ok=True)

def save_image(result, filename: str) -> None:
    """Save base64 image response to file."""
    image_base64 = result.data[0].b64_json
    with open(f"output_images/{filename}", "wb") as f:
        f.write(base64.b64decode(image_base64))

# Basic text-to-image generation (基本文生图生成)
result = client.images.generate(
    model="gpt-image-1.5",
    prompt="Your detailed prompt here",
    quality="high",  # or "low" for faster iteration (或"low"用于更快迭代)
    n=1  # number of variations (变体数量)
)

save_image(result, "output.png")

result = client.images.edit(
    model="gpt-image-1.5",
    input_fidelity="high",  # Essential for identity preservation (身份保留至关重要)
    quality="high",
    image=[
        open("input_images/source.png", "rb"),
        open("input_images/style_reference.png", "rb"),
    ],
    prompt="""
    Apply the artistic style from Image 2 to the subject in Image 1.
    (将图片2的艺术风格应用到图片1的主体上。)

    PRESERVE: subject's identity, pose, and composition
    (保留：主体的身份、姿势和构图)
    CHANGE: artistic style, color palette, texture treatment
    (改变：艺术风格、调色板、纹理处理)

    Do not add new elements. Maintain subject likeness exactly.
    (不要添加新元素。精确保持主体相似度。)
    """
)

save_image(result, "styled_output.png")