Formerly KrillinAI, VideoTranslation and Dubbing Tool for 120+ Languages. Open-source AI workspace for creators, powered by Codex. Video downloader, video translation, create videos, images, audios, and edits with Agents in one place. For your Social Media like YouTube, TikTok, Instagram, etc.
TypeScript agent-skillscodex-clideskop-appdubbingimage-generationlocalizationskillsttsvideo-editingvideo-generationvideo-transcriptionvideo-translation
fix(templates): normalize covers and localize prompts
卡皮巴拉f0b56a508b71
diff --git a/template/image-generation/acrylic-hand-painted-split-poster/1/cover.jpg b/template/image-generation/acrylic-hand-painted-split-poster/1/cover.jpgindex a6256931..00f712fc 100644Binary files a/template/image-generation/acrylic-hand-painted-split-poster/1/cover.jpg and b/template/image-generation/acrylic-hand-painted-split-poster/1/cover.jpg differdiff --git a/template/image-generation/animated-campus-world-reference-board/1/cover.jpg b/template/image-generation/animated-campus-world-reference-board/1/cover.jpgindex c6122113..fe11ffc4 100644Binary files a/template/image-generation/animated-campus-world-reference-board/1/cover.jpg and b/template/image-generation/animated-campus-world-reference-board/1/cover.jpg differdiff --git a/template/image-generation/architecture-photo-concept-sketch-diptych/1/cover.jpg b/template/image-generation/architecture-photo-concept-sketch-diptych/1/cover.jpgindex 7aae1d08..2ba65d82 100644Binary files a/template/image-generation/architecture-photo-concept-sketch-diptych/1/cover.jpg and b/template/image-generation/architecture-photo-concept-sketch-diptych/1/cover.jpg differdiff --git a/template/image-generation/bento-food-packaging-concept/1/cover.jpg b/template/image-generation/bento-food-packaging-concept/1/cover.jpgindex 6125da81..20612f79 100644Binary files a/template/image-generation/bento-food-packaging-concept/1/cover.jpg and b/template/image-generation/bento-food-packaging-concept/1/cover.jpg differdiff --git a/template/image-generation/berry-smoothie-commercial-photo/1/cover.jpg b/template/image-generation/berry-smoothie-commercial-photo/1/cover.jpgindex 1fb3e605..4cff6b13 100644Binary files a/template/image-generation/berry-smoothie-commercial-photo/1/cover.jpg and b/template/image-generation/berry-smoothie-commercial-photo/1/cover.jpg differdiff --git a/template/image-generation/beverage-food-marketing-hero/1/cover.jpg b/template/image-generation/beverage-food-marketing-hero/1/cover.jpgindex bd7c4e69..f3393046 100644Binary files a/template/image-generation/beverage-food-marketing-hero/1/cover.jpg and b/template/image-generation/beverage-food-marketing-hero/1/cover.jpg differdiff --git a/template/image-generation/blue-white-dunhuang-embroidery-quadriptych/1/cover.jpg b/template/image-generation/blue-white-dunhuang-embroidery-quadriptych/1/cover.jpgindex 77668bd0..50eca6a3 100644Binary files a/template/image-generation/blue-white-dunhuang-embroidery-quadriptych/1/cover.jpg and b/template/image-generation/blue-white-dunhuang-embroidery-quadriptych/1/cover.jpg differdiff --git a/template/image-generation/boutique-ecommerce-product-flatlay-grid/1/cover.jpg b/template/image-generation/boutique-ecommerce-product-flatlay-grid/1/cover.jpgindex 9b191173..40054dbc 100644Binary files a/template/image-generation/boutique-ecommerce-product-flatlay-grid/1/cover.jpg and b/template/image-generation/boutique-ecommerce-product-flatlay-grid/1/cover.jpg differdiff --git a/template/image-generation/chinese-luxury-beauty-poster/1/cover.jpg b/template/image-generation/chinese-luxury-beauty-poster/1/cover.jpgindex 4b344603..8699f2f0 100644Binary files a/template/image-generation/chinese-luxury-beauty-poster/1/cover.jpg and b/template/image-generation/chinese-luxury-beauty-poster/1/cover.jpg differdiff --git a/template/image-generation/cinematic-action-rpg-screenshot/1/cover.jpg b/template/image-generation/cinematic-action-rpg-screenshot/1/cover.jpgindex 83c66372..b8556a1d 100644Binary files a/template/image-generation/cinematic-action-rpg-screenshot/1/cover.jpg and b/template/image-generation/cinematic-action-rpg-screenshot/1/cover.jpg differdiff --git a/template/image-generation/city-corner-3d-billboard-photography/1/cover.jpg b/template/image-generation/city-corner-3d-billboard-photography/1/cover.jpgindex c0700d1c..68597883 100644Binary files a/template/image-generation/city-corner-3d-billboard-photography/1/cover.jpg and b/template/image-generation/city-corner-3d-billboard-photography/1/cover.jpg differdiff --git a/template/image-generation/concept-sofa-development-board/1/cover.jpg b/template/image-generation/concept-sofa-development-board/1/cover.jpgindex 534e7d64..d132d020 100644Binary files a/template/image-generation/concept-sofa-development-board/1/cover.jpg and b/template/image-generation/concept-sofa-development-board/1/cover.jpg differdiff --git a/template/image-generation/concept-sofa-development-board/1/template.json b/template/image-generation/concept-sofa-development-board/1/template.jsonindex 1d4c9938..e1758059 100644--- a/template/image-generation/concept-sofa-development-board/1/template.json+++ b/template/image-generation/concept-sofa-development-board/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Design Concept: The Crumple Chair Core Philosophy: Translating the \"controlled chaos\" of a tossed paper ball into a sculptural, high-comfort seating experience.\n\nStage 1: Observation & Morphological Analysis The goal is to deconstruct the image of the crumpled paper into usable geometric data. Crease Mapping: Identify the primary \"valley\" and \"ridge\" lines. These represent potential structural ribs or seams in the chair. Faceted Planes: Break down the sphere into a series of non-uniform polygons. Each flat surface of the paper becomes a potential panel for the chair’s upholstery or shell. Shadow Study: Analyze how the \"tossed\" form creates deep recesses. These natural pockets guide where the user’s weight will be cradled.\n\nStage 2: Iterative Form Exploration Moving from a sphere to a seat through \"Digital Crumpling.\" Subtractive Sculpting: Imagine the paper ball as a solid mass. Use Boolean operations to \"carve out\" a seating cavity that fits the human form while maintaining the external jagged texture. Tension Simulation: Use 3D software (like Rhino or Blender) to simulate a flat sheet of material being compressed. This ensures the folds look authentic and not \"modeled.\" The \"Toss\" Logic: Experiment with gravity-based simulation dropping a digital mesh to see how it settles naturally, mimicking the \"tossed\" origin.\n\nStage 3: Ergonomic Translation & Blueprinting Refining the raw aesthetic into a functional object. The Comfort Core: Overlay a standard ergonomic template (Seating Angle: 105°–110°) over the crumpled form. Adjust the internal \"folds\" to provide lumbar support and pressure relief. Blueprint Generation: Create technical orthographic views (Front, Side, Top). Map out the dimensions: Seat Height: 450mm Total Width: 850mm Surface Smoothing: Maintain the sharp \"paper edges\" on the exterior shell while softening the interior contact points for skin comfort.\n\nStage 4: Structural Integration & Scaling Making the concept physically viable. The Skeleton: Design a hidden internal frame (likely CNC-bent steel rods or a 3D-printed lattice) that follows the most prominent ridges of the paper folds to provide rigidity. Material Selection: * Option A (High-End): Faceted, cast aluminum with a white powder coat. Option B (Soft): Vacuum-formed recycled plastic shell covered in \"memory-fold\" technical fabric that retains a wrinkled appearance.\n\nStage 5: Final Prototyping & Material Finish Textural Replication: Apply a matte, slightly porous finish to the material to mimic the tactile feel of heavy-bond paper. Lighting Contrast: Use directional studio lighting in the final renders to emphasize the \"tossed\" shadows, making the chair look like a giant piece of discarded inspiration. Design Tip: To keep the \"tossed\" look authentic, avoid symmetry. The most compelling aspect of a crumpled paper ball is its unique irregularity—ensure the left and right sides of the chair are balance-equivalent but not identical"+ "prompt": "设计理念:Crumple Chair 核心理念:将扔纸球的“受控混乱”转化为雕塑般的高舒适度座椅体验。\n\n第一阶段:观察和形态分析目标是将皱巴巴的纸的图像解构为可用的几何数据。折痕映射:识别主要的“谷”和“脊”线。这些代表椅子中潜在的结构肋或接缝。多面平面:将球体分解为一系列不均匀的多边形。纸张的每个平坦表面都成为椅子内饰或外壳的潜在面板。阴影研究:分析“抛掷”形式如何产生深凹。这些天然的口袋引导用户的体重放置在哪里。\n\n第二阶段:迭代形式探索通过“数字压皱”从球体转移到座位。减法雕刻:将纸球想象成一个固体。使用布尔运算“雕刻”出适合人体形状的座位空腔,同时保持外部锯齿状纹理。张力模拟:使用 3D 软件(如 Rhino 或 Blender)来模拟被压缩的平板材料。这确保了褶皱看起来真实而不是“模仿”。 “抛掷”逻辑:尝试基于重力的模拟,让数字网格掉落,看看它如何自然沉降,模仿“抛掷”原点。\n\n第三阶段:人体工程学翻译和蓝图将原始美学提炼为功能性物体。舒适核心:在皱巴巴的形状上覆盖标准人体工学模板(座椅角度:105°–110°)。调整内部“褶皱”以提供腰部支撑和压力缓解。蓝图生成:创建技术正交视图(正面、侧面、顶部)。绘制尺寸: 座椅高度:450 毫米 总宽度:850 毫米 表面平滑:保持外壳上锋利的“纸边缘”,同时软化内部接触点,使皮肤舒适。\n\n第四阶段:结构整合和扩展使概念在物理上可行。骨架:设计一个隐藏的内部框架(可能是 CNC 弯曲钢棒或 3D 打印的格子),该框架遵循纸折最突出的脊线以提供刚性。材料选择: * 选项 A(高端):带白色粉末涂层的多面铸铝。选项 B(软质):真空成型再生塑料外壳,覆盖“记忆折叠”技术织物,保留褶皱外观。\n\n第 5 阶段:最终原型制作和材料饰面纹理复制:在材料上应用哑光、微孔饰面,以模仿重磅纸的触感。照明对比:在最终渲染中使用定向工作室照明来强调“抛掷”阴影,使椅子看起来像是一块被丢弃的灵感的巨大碎片。设计提示:为了保持“抛掷”的外观真实,避免对称。皱巴巴的纸球最引人注目的方面是其独特的不规则性——确保椅子的左右两侧保持视觉平衡,但不要完全相同"},"en-US": {"prompt": "Design Concept: The Crumple Chair Core Philosophy: Translating the \"controlled chaos\" of a tossed paper ball into a sculptural, high-comfort seating experience.\n\nStage 1: Observation & Morphological Analysis The goal is to deconstruct the image of the crumpled paper into usable geometric data. Crease Mapping: Identify the primary \"valley\" and \"ridge\" lines. These represent potential structural ribs or seams in the chair. Faceted Planes: Break down the sphere into a series of non-uniform polygons. Each flat surface of the paper becomes a potential panel for the chair’s upholstery or shell. Shadow Study: Analyze how the \"tossed\" form creates deep recesses. These natural pockets guide where the user’s weight will be cradled.\n\nStage 2: Iterative Form Exploration Moving from a sphere to a seat through \"Digital Crumpling.\" Subtractive Sculpting: Imagine the paper ball as a solid mass. Use Boolean operations to \"carve out\" a seating cavity that fits the human form while maintaining the external jagged texture. Tension Simulation: Use 3D software (like Rhino or Blender) to simulate a flat sheet of material being compressed. This ensures the folds look authentic and not \"modeled.\" The \"Toss\" Logic: Experiment with gravity-based simulation dropping a digital mesh to see how it settles naturally, mimicking the \"tossed\" origin.\n\nStage 3: Ergonomic Translation & Blueprinting Refining the raw aesthetic into a functional object. The Comfort Core: Overlay a standard ergonomic template (Seating Angle: 105°–110°) over the crumpled form. Adjust the internal \"folds\" to provide lumbar support and pressure relief. Blueprint Generation: Create technical orthographic views (Front, Side, Top). Map out the dimensions: Seat Height: 450mm Total Width: 850mm Surface Smoothing: Maintain the sharp \"paper edges\" on the exterior shell while softening the interior contact points for skin comfort.\n\nStage 4: Structural Integration & Scaling Making the concept physically viable. The Skeleton: Design a hidden internal frame (likely CNC-bent steel rods or a 3D-printed lattice) that follows the most prominent ridges of the paper folds to provide rigidity. Material Selection: * Option A (High-End): Faceted, cast aluminum with a white powder coat. Option B (Soft): Vacuum-formed recycled plastic shell covered in \"memory-fold\" technical fabric that retains a wrinkled appearance.\n\nStage 5: Final Prototyping & Material Finish Textural Replication: Apply a matte, slightly porous finish to the material to mimic the tactile feel of heavy-bond paper. Lighting Contrast: Use directional studio lighting in the final renders to emphasize the \"tossed\" shadows, making the chair look like a giant piece of discarded inspiration. Design Tip: To keep the \"tossed\" look authentic, avoid symmetry. The most compelling aspect of a crumpled paper ball is its unique irregularity—ensure the left and right sides of the chair are balance-equivalent but not identical"diff --git a/template/image-generation/consistent-interior-multiview-grid/1/cover.jpg b/template/image-generation/consistent-interior-multiview-grid/1/cover.jpgindex 8c0bf35b..7c613e67 100644Binary files a/template/image-generation/consistent-interior-multiview-grid/1/cover.jpg and b/template/image-generation/consistent-interior-multiview-grid/1/cover.jpg differdiff --git a/template/image-generation/cooking-sketch-illustration-story/1/cover.jpg b/template/image-generation/cooking-sketch-illustration-story/1/cover.jpgindex 6cf7cfdf..98c1e8fa 100644Binary files a/template/image-generation/cooking-sketch-illustration-story/1/cover.jpg and b/template/image-generation/cooking-sketch-illustration-story/1/cover.jpg differdiff --git a/template/image-generation/croissant-baking-storyboard/1/cover.jpg b/template/image-generation/croissant-baking-storyboard/1/cover.jpgindex 78feafb8..1aafbfcc 100644Binary files a/template/image-generation/croissant-baking-storyboard/1/cover.jpg and b/template/image-generation/croissant-baking-storyboard/1/cover.jpg differdiff --git a/template/image-generation/croissant-baking-storyboard/1/template.json b/template/image-generation/croissant-baking-storyboard/1/template.jsonindex d09990de..922e7b64 100644--- a/template/image-generation/croissant-baking-storyboard/1/template.json+++ b/template/image-generation/croissant-baking-storyboard/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Create a crisp, clean infographic storyboard poster for THE CROISSANT BAKER. Wide 16:9 layout, white background, black borders, bold black typography, premium Pixar 3D stylized rendering, bright vivid colors — warm golden yellows, rich buttery creams, flaky browns, soft pastry whites, warm French bakery morning light.\nTop header:\n\nTHE CROISSANT BAKER\nTOTAL VIDEO TIME: 12 SECONDS\n8 SHOTS · WARM · FLAKY · IRRESISTIBLE\nLegend icons: ACTION, HEAT, TIME HINT, INGREDIENT\nThin warm golden accent line running full width beneath header\n\nSame Pixar-style young French male baker throughout: white baker's jacket, flour-dusted hands, warm authentic French boulangerie setting, marble countertop, warm morning light streaming through windows, bread racks in background. Bright, warm, delicious. Every panel a completely different composition and color.\n8 panels:\n\nTHE OPENER — Wide shot of baker arriving at the boulangerie before dawn, tying apron, switching on the warm kitchen lights, marble counter visible, bread racks behind, flour dusting the air, full world established, bright and cinematic\nTHE BUTTER BLOCK — Baker slams a massive cold block of European butter onto the marble counter with both hands, dramatic impact, flour cloud puffing up, close-up on hands, this is the bones moment — the start of everything\nTHE LAMINATION — Baker folding the dough over the butter block precisely, rolling pin pressing down hard, layers building, side angle shot showing the beautiful layering beginning, confident and skilled\nTHE ROLL — Dough rolled out into a large thin sheet, baker leaning into the rolling pin with full body weight, marble counter, flour dusting everywhere, wide shot showing the scale of the dough\nTHE SHAPE — Triangles cut from the dough, baker rolling each one from the wide end into a tight crescent, hands moving fast and confident, close-up on the shaping, beautiful and precise\nTHE EGG WASH — Baker brushing golden egg wash over each shaped croissant with a pastry brush, each one glistening beautifully, close-up overhead angle, warm golden color, stunning composition\nTHE OVEN — Croissants slid into the blazing hot oven on a tray, oven door closed, through the oven glass croissants visibly puffing and turning deep golden, layers separating dramatically, warm orange glow\nTHE TEAR — Baker pulls a perfect golden croissant from the rack, holds it up, tears it open slowly revealing hundreds of impossibly flaky buttery layers inside, steam escaping, butter glistening — this is the cheese pull moment, the hero shot of the entire video\n\nFooter:\n\nVIDEO FLOW: 8 shots × 1.5s = 12 seconds. Butter block to the tear.\nCAMERA TIPS: wide on opener, close-up on butter slam and shaping, side angle on lamination, overhead on egg wash, oven glass for panel 7, extreme close-up on the tear reveal\nLIGHT & STYLE: warm golden French bakery morning light, buttery cream tones, flour dust in the air, bright vivid Pixar colors, shallow depth of field on close-ups\nBAKER NOTES: one baker, one perfect croissant, one irresistible tear. The lamination layers and the final tear are everything — make them stunning."+ "prompt": "为《可颂面包师》制作一张清晰、干净的信息图表故事板海报。宽屏 16:9 布局、白色背景、黑色边框、大胆的黑色排版、优质皮克斯 3D 风格化渲染、明亮生动的色彩 — 温暖的金黄色、浓郁的奶油色、酥皮棕色、柔和的面点白色、温暖的法式面包店晨光。\n顶部标题:\n\n牛角面包师\n视频总时间:12 秒\n8 个镜头 · 温暖 · 酥脆 · 不可抗拒\n图例图标:动作、加热、时间提示、原料\n细而温暖的金色强调线在标题下方全宽延伸\n\n皮克斯风格的年轻法国男面包师贯穿始终:白色烘焙师制服,沾满面粉的手,温暖的正宗法式面包房环境,大理石台面,温暖的晨光透过窗户,背景是面包架。明亮、温暖、美味。每个面板都有完全不同的构图和色彩。\n8 个面板:\n\n开场——面包师在黎明前到达面包店的广角镜头,系上围裙,打开温暖的厨房灯光,大理石柜台可见,面包架在后面,面粉撒在空气中,完整的世界已建立,明亮而电影化\n黄油块——面包师用双手将一块巨大的冷欧洲黄油猛击到大理石柜台上,戏剧性的撞击,面粉云膨胀起来,手上的特写,这是奠定基础的关键时刻——一切的开始\n开酥—— 面包师将面团精确地折叠在黄油块上,擀面杖用力压下,分层,侧角拍摄显示美丽的分层开始,自信而熟练\n擀面—— 面团被擀成一张大薄片,面包师以全身重量靠在擀面杖上,大理石柜台,面粉撒满各处,广角镜头显示面团的大小\n整形——从面团上切出三角形,面包师将每个三角形从宽端卷成紧密的新月形,双手快速而自信地移动,特写造型,美丽而精确\n刷蛋液—— 面包师用糕点刷在每个形状的牛角面包上刷上金色的鸡蛋洗,每个牛角面包都闪闪发光,特写头顶角度,温暖的金色,令人惊叹的构图\n烤箱——羊角面包放在托盘上滑入炽热的烤箱,烤箱门关闭,透过烤箱玻璃,羊角面包明显膨化并变成深金色,层层明显分离,发出温暖的橙色光芒\n撕开——面包师从架子上拿出一个完美的金色牛角面包,举起它,慢慢地把它撕开,露出里面数百层令人惊叹的酥脆黄油层次,蒸汽逸出,黄油闪闪发光——这是如同拉丝镜头般诱人的关键时刻,是整个视频的英雄镜头\n\n页脚:\n\n视频流程:8 个镜头 × 1.5 秒 = 12 秒。从黄油块到最终撕开。\n相机提示:开场使用广角,黄油猛击和成型的特写,开酥使用侧面角度,刷蛋液使用俯拍,面板 7 的烤箱玻璃,撕开呈现使用极近特写\n灯光和风格:温暖的金色法式面包店晨光、黄油奶油色调、空气中的面粉灰尘、明亮生动的皮克斯色彩、特写镜头的浅景深\n面包师笔记:一位面包师,一份完美的羊角面包,一次令人无法抗拒的撕开呈现。开酥层次和最后的撕裂就是一切——让它们令人惊叹。"},"en-US": {"prompt": "Create a crisp, clean infographic storyboard poster for THE CROISSANT BAKER. Wide 16:9 layout, white background, black borders, bold black typography, premium Pixar 3D stylized rendering, bright vivid colors — warm golden yellows, rich buttery creams, flaky browns, soft pastry whites, warm French bakery morning light.\nTop header:\n\nTHE CROISSANT BAKER\nTOTAL VIDEO TIME: 12 SECONDS\n8 SHOTS · WARM · FLAKY · IRRESISTIBLE\nLegend icons: ACTION, HEAT, TIME HINT, INGREDIENT\nThin warm golden accent line running full width beneath header\n\nSame Pixar-style young French male baker throughout: white baker's jacket, flour-dusted hands, warm authentic French boulangerie setting, marble countertop, warm morning light streaming through windows, bread racks in background. Bright, warm, delicious. Every panel a completely different composition and color.\n8 panels:\n\nTHE OPENER — Wide shot of baker arriving at the boulangerie before dawn, tying apron, switching on the warm kitchen lights, marble counter visible, bread racks behind, flour dusting the air, full world established, bright and cinematic\nTHE BUTTER BLOCK — Baker slams a massive cold block of European butter onto the marble counter with both hands, dramatic impact, flour cloud puffing up, close-up on hands, this is the bones moment — the start of everything\nTHE LAMINATION — Baker folding the dough over the butter block precisely, rolling pin pressing down hard, layers building, side angle shot showing the beautiful layering beginning, confident and skilled\nTHE ROLL — Dough rolled out into a large thin sheet, baker leaning into the rolling pin with full body weight, marble counter, flour dusting everywhere, wide shot showing the scale of the dough\nTHE SHAPE — Triangles cut from the dough, baker rolling each one from the wide end into a tight crescent, hands moving fast and confident, close-up on the shaping, beautiful and precise\nTHE EGG WASH — Baker brushing golden egg wash over each shaped croissant with a pastry brush, each one glistening beautifully, close-up overhead angle, warm golden color, stunning composition\nTHE OVEN — Croissants slid into the blazing hot oven on a tray, oven door closed, through the oven glass croissants visibly puffing and turning deep golden, layers separating dramatically, warm orange glow\nTHE TEAR — Baker pulls a perfect golden croissant from the rack, holds it up, tears it open slowly revealing hundreds of impossibly flaky buttery layers inside, steam escaping, butter glistening — this is the cheese pull moment, the hero shot of the entire video\n\nFooter:\n\nVIDEO FLOW: 8 shots × 1.5s = 12 seconds. Butter block to the tear.\nCAMERA TIPS: wide on opener, close-up on butter slam and shaping, side angle on lamination, overhead on egg wash, oven glass for panel 7, extreme close-up on the tear reveal\nLIGHT & STYLE: warm golden French bakery morning light, buttery cream tones, flour dust in the air, bright vivid Pixar colors, shallow depth of field on close-ups\nBAKER NOTES: one baker, one perfect croissant, one irresistible tear. The lamination layers and the final tear are everything — make them stunning."diff --git a/template/image-generation/deep-sea-freediving-adventure-photo/1/cover.jpg b/template/image-generation/deep-sea-freediving-adventure-photo/1/cover.jpgindex 56751a88..549de212 100644Binary files a/template/image-generation/deep-sea-freediving-adventure-photo/1/cover.jpg and b/template/image-generation/deep-sea-freediving-adventure-photo/1/cover.jpg differdiff --git a/template/image-generation/desert-car-wash-documentary-photo/1/cover.jpg b/template/image-generation/desert-car-wash-documentary-photo/1/cover.jpgindex 43af7a22..181803a5 100644Binary files a/template/image-generation/desert-car-wash-documentary-photo/1/cover.jpg and b/template/image-generation/desert-car-wash-documentary-photo/1/cover.jpg differdiff --git a/template/image-generation/duotone-studio-commercial-portrait/1/cover.jpg b/template/image-generation/duotone-studio-commercial-portrait/1/cover.jpgindex caf1fc7e..a6bd4559 100644Binary files a/template/image-generation/duotone-studio-commercial-portrait/1/cover.jpg and b/template/image-generation/duotone-studio-commercial-portrait/1/cover.jpg differdiff --git a/template/image-generation/exploded-food-infographic/1/cover.jpg b/template/image-generation/exploded-food-infographic/1/cover.jpgindex 25321636..a914a16f 100644Binary files a/template/image-generation/exploded-food-infographic/1/cover.jpg and b/template/image-generation/exploded-food-infographic/1/cover.jpg differdiff --git a/template/image-generation/exploded-food-infographic/1/template.json b/template/image-generation/exploded-food-infographic/1/template.jsonindex 4ae60e8a..8975b3a8 100644--- a/template/image-generation/exploded-food-infographic/1/template.json+++ b/template/image-generation/exploded-food-infographic/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Create a hyper-realistic exploded vertical infographic composition of tacos.\n\nTop → Bottom structure:\nFresh Lettuce (crisp green texture with natural folds)\n→ Tomato & Salsa Layer (juicy diced tomatoes and salsa mix)\n→ Melted Cheese (smooth cheddar texture)\n→ Grilled Meat Filling (juicy seasoned meat detail)\n→ Taco Shell Base (crispy golden shell texture)\nPerfect vertical alignment, rustic background, soft studio lighting, realistic shadows beneath each floating element.\n\nAdd clean infographic text labels with thin pointer lines using these exact labels:\n“Lettuce”\n“Salsa”\n“Cheese”\n“Meat”\n“Shell”\nUltra-detailed food textures, premium commercial aesthetic, 8K."+ "prompt": "创建超写实的塔可(Taco)爆炸分解式竖版信息图。\n\n顶部→底部结构:\n新鲜生菜(脆嫩的绿色质地,自然褶皱)\n→ 番茄和莎莎酱层(多汁的番茄丁和莎莎酱混合)\n→ 融化的奶酪(光滑的切达干酪质地)\n→ 烤肉馅(多汁的调味肉细节)\n→ Taco Shell Base(脆皮金壳质地)\n完美的垂直对齐、质朴的背景、柔和的工作室灯光、每个浮动元素下方的真实阴影。\n\n使用这些精确的标签添加带有细指针线的干净的信息图表文本标签:\n“生菜”\n“莎莎酱”\n“奶酪”\n“肉”\n“塔可脆壳”\n超详细的食物纹理,优质的商业美学,8K。"},"en-US": {"prompt": "Create a hyper-realistic exploded vertical infographic composition of tacos.\n\nTop → Bottom structure:\nFresh Lettuce (crisp green texture with natural folds)\n→ Tomato & Salsa Layer (juicy diced tomatoes and salsa mix)\n→ Melted Cheese (smooth cheddar texture)\n→ Grilled Meat Filling (juicy seasoned meat detail)\n→ Taco Shell Base (crispy golden shell texture)\nPerfect vertical alignment, rustic background, soft studio lighting, realistic shadows beneath each floating element.\n\nAdd clean infographic text labels with thin pointer lines using these exact labels:\n“Lettuce”\n“Salsa”\n“Cheese”\n“Meat”\n“Shell”\nUltra-detailed food textures, premium commercial aesthetic, 8K."diff --git a/template/image-generation/felt-country-miniature-world/1/cover.jpg b/template/image-generation/felt-country-miniature-world/1/cover.jpgindex fd58144c..4ae6e388 100644Binary files a/template/image-generation/felt-country-miniature-world/1/cover.jpg and b/template/image-generation/felt-country-miniature-world/1/cover.jpg differdiff --git a/template/image-generation/felt-country-miniature-world/1/template.json b/template/image-generation/felt-country-miniature-world/1/template.jsonindex 49cbb3da..8a8fb557 100644--- a/template/image-generation/felt-country-miniature-world/1/template.json+++ b/template/image-generation/felt-country-miniature-world/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Country: [INSERT COUNTRY NAME]\n\nA miniature felt diorama made of fluffy yarn, wool, and needlework, designed as a single cohesive small world (not a collage), reflecting the real landscape, culture, and daily life of the country.\n\nStructure:\n- Select 4–7 naturally integrated elements (landscape, architecture, transport, street life, culture)\n- Everything must flow together as one environment\n\nForeground:\nEveryday life — small shops, cafes, markets, people interacting, local clothing, street details, subtle movement, warm human presence\n\nMidground:\nConnecting flow — streets, bridges, rivers, paths, transport, cultural spaces that guide the eye naturally from front to back\n\nBackground:\nOne strong identity anchor — landmark, skyline, mountain, or symbolic landscape clearly representing the country (keep it clean, not overcrowded)\n\nStyle:\nHandmade wool felt, yarn, needle-felt textures, visible fibers, soft edges, miniature craftsmanship, premium diorama realism\n\nLighting:\nWarm golden hour or bright noon, soft shadows, clear visibility, gentle glow enhancing depth\n\nColor:\nSoft but saturated palette reflecting the country’s natural tones (greens, sky blues, architectural hues, cultural accents), warm and balanced\n\nComposition:\nVertical frame, balanced or centered perspective, slight top-down or immersive angle, clear foreground–midground–background depth\n\nMood:\nWarm, emotional, calm, refined — like a handcrafted fairy-tale version of real life (not childish)\n\nRules:\n- No logos or text overlays\n- No collage-style composition\n- No random symbolic clutter\n- Keep it realistic but stylized as a handmade miniature world\n\nQuality:\n16K, ultra-detailed, hyper-realistic miniature textures, cinematic depth, sharp focus\n\noutput_goal:\nA single, cohesive felt diorama world that instantly conveys the identity and atmosphere of the chosen country through integrated landscape, culture, and daily life"+ "prompt": "国家/地区:[插入国家/地区名称]\n\n由蓬松的纱线、羊毛和针线制成的微型毛毡立体模型,设计成一个单一的有凝聚力的小世界(不是拼贴画),反映了这个国家的真实景观、文化和日常生活。\n\n结构:\n- 选择 4-7 个自然整合的元素(景观、建筑、交通、街道生活、文化)\n- 一切都必须作为一个环境一起流动\n\n前景:\n日常生活——小商店、咖啡馆、市场、人们互动、当地服装、街道细节、微妙的动作、温暖的人类存在\n\n中景:\n连接流线——街道、桥梁、河流、小路、交通、文化空间,自然地引导视线从前到后\n\n背景:\n一个强有力的身份锚点——地标、天际线、山脉或象征性景观,清楚地代表国家(保持干净,不要过度拥挤)\n\n风格:\n手工羊毛毡、纱线、针毡纹理、可见纤维、柔软边缘、微型工艺、高级立体模型现实主义\n\n照明:\n温暖的黄金时段或明亮的正午,柔和的阴影,清晰的可见度,柔和的辉光增强深度\n\n颜色:\n柔和但饱和的调色板反映了该国的自然色调(绿色、天蓝色、建筑色调、文化特色)、温暖而平衡\n\n成分:\n垂直框架,平衡或居中视角,轻微自上而下或沉浸式角度,清晰的前景-中景-背景深度\n\n心情:\n温暖、感性、平静、精致——就像现实生活的手工童话版本(不幼稚)\n\n规则:\n- 无徽标或文字覆盖\n- 没有拼贴风格的构图\n- 没有随机的符号混乱\n- 保持现实,但风格化为手工制作的微型世界\n\n质量:\n16K、超细致、超现实的微型纹理、电影深度、锐利的焦点\n\n输出目标:\n一个单一、有凝聚力的毛毡立体模型世界,通过整合的景观、文化和日常生活立即传达所选国家的身份和氛围"},"en-US": {"prompt": "Country: [INSERT COUNTRY NAME]\n\nA miniature felt diorama made of fluffy yarn, wool, and needlework, designed as a single cohesive small world (not a collage), reflecting the real landscape, culture, and daily life of the country.\n\nStructure:\n- Select 4–7 naturally integrated elements (landscape, architecture, transport, street life, culture)\n- Everything must flow together as one environment\n\nForeground:\nEveryday life — small shops, cafes, markets, people interacting, local clothing, street details, subtle movement, warm human presence\n\nMidground:\nConnecting flow — streets, bridges, rivers, paths, transport, cultural spaces that guide the eye naturally from front to back\n\nBackground:\nOne strong identity anchor — landmark, skyline, mountain, or symbolic landscape clearly representing the country (keep it clean, not overcrowded)\n\nStyle:\nHandmade wool felt, yarn, needle-felt textures, visible fibers, soft edges, miniature craftsmanship, premium diorama realism\n\nLighting:\nWarm golden hour or bright noon, soft shadows, clear visibility, gentle glow enhancing depth\n\nColor:\nSoft but saturated palette reflecting the country’s natural tones (greens, sky blues, architectural hues, cultural accents), warm and balanced\n\nComposition:\nVertical frame, balanced or centered perspective, slight top-down or immersive angle, clear foreground–midground–background depth\n\nMood:\nWarm, emotional, calm, refined — like a handcrafted fairy-tale version of real life (not childish)\n\nRules:\n- No logos or text overlays\n- No collage-style composition\n- No random symbolic clutter\n- Keep it realistic but stylized as a handmade miniature world\n\nQuality:\n16K, ultra-detailed, hyper-realistic miniature textures, cinematic depth, sharp focus\n\noutput_goal:\nA single, cohesive felt diorama world that instantly conveys the identity and atmosphere of the chosen country through integrated landscape, culture, and daily life"diff --git a/template/image-generation/fifteen-step-photorealistic-recipe/1/cover.jpg b/template/image-generation/fifteen-step-photorealistic-recipe/1/cover.jpgindex 7a9e6dc9..64361abf 100644Binary files a/template/image-generation/fifteen-step-photorealistic-recipe/1/cover.jpg and b/template/image-generation/fifteen-step-photorealistic-recipe/1/cover.jpg differdiff --git a/template/image-generation/five-panel-cafe-character-collage/1/cover.jpg b/template/image-generation/five-panel-cafe-character-collage/1/cover.jpgindex 7c235025..124d67fc 100644Binary files a/template/image-generation/five-panel-cafe-character-collage/1/cover.jpg and b/template/image-generation/five-panel-cafe-character-collage/1/cover.jpg differdiff --git a/template/image-generation/football-player-stats-doodle-poster/1/cover.jpg b/template/image-generation/football-player-stats-doodle-poster/1/cover.jpgindex 25071b89..52d70edb 100644Binary files a/template/image-generation/football-player-stats-doodle-poster/1/cover.jpg and b/template/image-generation/football-player-stats-doodle-poster/1/cover.jpg differdiff --git a/template/image-generation/football-player-stats-doodle-poster/1/template.json b/template/image-generation/football-player-stats-doodle-poster/1/template.jsonindex 7384f319..27eb93ad 100644--- a/template/image-generation/football-player-stats-doodle-poster/1/template.json+++ b/template/image-generation/football-player-stats-doodle-poster/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Create a scrapbook doodle-style football poster of [PLAYER_NAME].\n\nAI auto-generates realistic career stats (club and national team).\n\nMain photo: realistic, unchanged ([PLAYER_NAME] in action or iconic pose).\n\nDoodles and handwritten notes: white and [PRIMARY_COLOR] neon ink (no warm colors unless specified), arrows, stars, scribbles, sketchy outlines. Add a glowing [PRIMARY_COLOR] outline around the player's body.\n\nLayout (bold handwritten titles and stats):\n- Top Title: \"[PLAYER_NAME]\" plus \"[NICKNAME/TITLE]\"\n- Club Career: total matches and goals plus 2–4 main clubs (apps and goals)\n- Highlights: 2–3 standout achievements (e.g., UCL goals, Ballon d'Or, top scorer)\n- National Team: caps and goals plus 1–2 international achievements\n\nStyle: modern football poster x notebook aesthetic, clean but energetic, slightly messy doodles, high contrast, neon accents on dark or muted background.\n\nImportant: all stats must be realistic and proportional to the player's real career."+ "prompt": "为 [PLAYER_NAME] 制作剪贴簿涂鸦风格的足球海报。\n\nAI 自动生成真实可信的职业生涯数据(俱乐部和国家队)。\n\n主照片:保持写实且不作改动([PLAYER_NAME]的动作或标志性姿势)。\n\n涂鸦和手写笔记:白色和 [PRIMARY_COLOR] 霓虹灯墨水(除非指定,否则没有暖色)、箭头、星星、涂鸦、粗略轮廓。在球员身体周围添加发光的 [PRIMARY_COLOR] 轮廓。\n\n布局(粗体手写标题和统计数据):\n- 顶部标题:“[PLAYER_NAME]”加“[NICKNAME/TITLE]”\n- 俱乐部生涯:总比赛数和进球数加上 2-4 个主要俱乐部(出场次数和进球数)\n- 亮点:2–3 项杰出成就(例如 UCL 进球、金球奖、最佳射手)\n- 国家队:出场次数和进球数以及 1-2 项国际成就\n\n风格:现代足球海报 x 笔记本美学,干净但充满活力,略显凌乱的涂鸦,高对比度,深色或柔和背景上的霓虹灯装饰。\n\n重要提示:所有统计数据都必须真实且与球员的真实职业生涯成比例。"},"en-US": {"prompt": "Create a scrapbook doodle-style football poster of [PLAYER_NAME].\n\nAI auto-generates realistic career stats (club and national team).\n\nMain photo: realistic, unchanged ([PLAYER_NAME] in action or iconic pose).\n\nDoodles and handwritten notes: white and [PRIMARY_COLOR] neon ink (no warm colors unless specified), arrows, stars, scribbles, sketchy outlines. Add a glowing [PRIMARY_COLOR] outline around the player's body.\n\nLayout (bold handwritten titles and stats):\n- Top Title: \"[PLAYER_NAME]\" plus \"[NICKNAME/TITLE]\"\n- Club Career: total matches and goals plus 2–4 main clubs (apps and goals)\n- Highlights: 2–3 standout achievements (e.g., UCL goals, Ballon d'Or, top scorer)\n- National Team: caps and goals plus 1–2 international achievements\n\nStyle: modern football poster x notebook aesthetic, clean but energetic, slightly messy doodles, high contrast, neon accents on dark or muted background.\n\nImportant: all stats must be realistic and proportional to the player's real career."diff --git a/template/image-generation/four-city-minimal-travel-poster-series/1/cover.jpg b/template/image-generation/four-city-minimal-travel-poster-series/1/cover.jpgindex 0763189d..0d325faa 100644Binary files a/template/image-generation/four-city-minimal-travel-poster-series/1/cover.jpg and b/template/image-generation/four-city-minimal-travel-poster-series/1/cover.jpg differdiff --git a/template/image-generation/four-city-minimal-travel-poster-series/1/template.json b/template/image-generation/four-city-minimal-travel-poster-series/1/template.jsonindex d1e7bcab..de42b9fe 100644--- a/template/image-generation/four-city-minimal-travel-poster-series/1/template.json+++ b/template/image-generation/four-city-minimal-travel-poster-series/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Minimalist flat travel poster illustration series of iconic destinations around the world, clean vector art style, Scandinavian color palette, soft pastel tones, calm atmospheric scenery, Hallstatt Austria lakeside village with alpine mountains and reflections, Bergen Norway colorful harbor houses with fjord background, Kyoto Japan traditional riverside town with cherry blossoms and pagoda, Santorini Greece white architecture with blue domes overlooking the sea, ultra clean composition, geometric landscape shapes, smooth gradients, elegant typography with destination names, modern tourism poster aesthetic, serene travel mood, crisp vector lines, cinematic wide-angle scenery, peaceful skies with stylized clouds, highly detailed environment art, contemporary flat illustration, premium editorial travel design, balanced composition, minimal shadows, dreamy vacation atmosphere, soft morning lighting, high-end Airbnb poster vibe, Behance trending style, 4k, ultra detailed"+ "prompt": "世界各地标志性目的地的极简平面旅行海报插画系列,干净的矢量艺术风格,斯堪的纳维亚调色板,柔和的色调,平静大气的风景,哈尔施塔特奥地利湖畔村庄与高山山脉和倒影,挪威卑尔根色彩缤纷的港口房屋与峡湾背景,京都日本传统河滨小镇与樱花和宝塔,圣托里尼希腊白色建筑与蓝色圆顶俯瞰大海,超干净的构图,几何景观形状,平滑的渐变,带有目的地名称的优雅排版,现代旅游海报美学,宁静的旅行心情,清晰的矢量线,电影广角风景,宁静的天空与风格化的云彩,高度详细的环境艺术,当代平面插画,优质编辑旅行设计,平衡的构图,最小的阴影,梦幻般的度假氛围,柔和的早晨灯光,高端 Airbnb 海报氛围,Behance 趋势风格,4k,超详细"},"en-US": {"prompt": "Minimalist flat travel poster illustration series of iconic destinations around the world, clean vector art style, Scandinavian color palette, soft pastel tones, calm atmospheric scenery, Hallstatt Austria lakeside village with alpine mountains and reflections, Bergen Norway colorful harbor houses with fjord background, Kyoto Japan traditional riverside town with cherry blossoms and pagoda, Santorini Greece white architecture with blue domes overlooking the sea, ultra clean composition, geometric landscape shapes, smooth gradients, elegant typography with destination names, modern tourism poster aesthetic, serene travel mood, crisp vector lines, cinematic wide-angle scenery, peaceful skies with stylized clouds, highly detailed environment art, contemporary flat illustration, premium editorial travel design, balanced composition, minimal shadows, dreamy vacation atmosphere, soft morning lighting, high-end Airbnb poster vibe, Behance trending style, 4k, ultra detailed"diff --git a/template/image-generation/four-panel-milkshake-recipe/1/cover.jpg b/template/image-generation/four-panel-milkshake-recipe/1/cover.jpgindex 0f1af638..d0e58926 100644Binary files a/template/image-generation/four-panel-milkshake-recipe/1/cover.jpg and b/template/image-generation/four-panel-milkshake-recipe/1/cover.jpg differdiff --git a/template/image-generation/french-new-wave-torn-paper-poster/1/cover.jpg b/template/image-generation/french-new-wave-torn-paper-poster/1/cover.jpgindex 36310bbe..f1af9b4a 100644Binary files a/template/image-generation/french-new-wave-torn-paper-poster/1/cover.jpg and b/template/image-generation/french-new-wave-torn-paper-poster/1/cover.jpg differdiff --git a/template/image-generation/french-new-wave-torn-paper-poster/1/template.json b/template/image-generation/french-new-wave-torn-paper-poster/1/template.jsonindex 59fa49c0..2a00a9ce 100644--- a/template/image-generation/french-new-wave-torn-paper-poster/1/template.json+++ b/template/image-generation/french-new-wave-torn-paper-poster/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Create a vertical poster composition on aged cream paper with a handmade analog feel. Use rough ripped paper edges, layered magazine cutouts, photocopy grain, halftone texture, ink bleed, and slightly imperfect screen-print registration. Keep the subject as the main black-and-white photographic portrait, placed prominently in the center or upper center. Surround the subject with graphic blocks of deep red, cobalt blue, warm yellow, black, and ivory.\n\nAdd supporting collage fragments such as a rainy European street, film-strip borders, newspaper clippings, urban silhouettes, and cinematic details, arranged like a handmade 1960s art-house movie poster. Use bold condensed typography with a strong visual hierarchy. Add large headline text: \"[MAIN TITLE]\". Add smaller subtitle text: \"[SUBTITLE]\". Add bottom text: \"[BRAND NAME / EVENT NAME / COMING SOON / DATE]\". If needed, include a small top line reading \"[TAGLINE]\".\n\nThe final result should feel cinematic, intellectual, rebellious, and editorial — like a lost 1960s European film poster with a strong point of view. Keep it raw, tactile, printed, imperfect, and handmade. Avoid a glossy modern finish."+ "prompt": "在老化的奶油色纸上创建具有手工模拟感觉的垂直海报构图。使用粗糙的撕裂纸边缘、分层的杂志剪纸、复印颗粒、半色调纹理、渗墨和稍微不完美的丝网印刷套准。将拍摄对象作为主要的黑白摄影肖像,突出地放置在中心或上部中心。用深红色、钴蓝色、暖黄色、黑色和象牙色的图形块围绕主题。\n\n添加辅助拼贴片段,例如雨天的欧洲街道、胶片边框、剪报、城市剪影和电影细节,排列得像手工制作的 20 世纪 60 年代艺术电影海报一样。使用具有强烈视觉层次结构的大胆压缩版式。添加大标题文本:“[MAIN TITLE]”。添加较小的字幕文本:“[SUBTITLE]”。添加底部文字:“[品牌名称/活动名称/即将推出/日期]”。如果需要,请在顶行添加一小行“[TAGLINE]”。\n\n最终的结果应该给人电影感、知性、叛逆和社论的感觉——就像一张丢失的 20 世纪 60 年代欧洲电影海报,具有强烈的观点。保持原始、触感、印刷、不完美和手工制作。避免光滑的现代饰面。"},"en-US": {"prompt": "Create a vertical poster composition on aged cream paper with a handmade analog feel. Use rough ripped paper edges, layered magazine cutouts, photocopy grain, halftone texture, ink bleed, and slightly imperfect screen-print registration. Keep the subject as the main black-and-white photographic portrait, placed prominently in the center or upper center. Surround the subject with graphic blocks of deep red, cobalt blue, warm yellow, black, and ivory.\n\nAdd supporting collage fragments such as a rainy European street, film-strip borders, newspaper clippings, urban silhouettes, and cinematic details, arranged like a handmade 1960s art-house movie poster. Use bold condensed typography with a strong visual hierarchy. Add large headline text: \"[MAIN TITLE]\". Add smaller subtitle text: \"[SUBTITLE]\". Add bottom text: \"[BRAND NAME / EVENT NAME / COMING SOON / DATE]\". If needed, include a small top line reading \"[TAGLINE]\".\n\nThe final result should feel cinematic, intellectual, rebellious, and editorial — like a lost 1960s European film poster with a strong point of view. Keep it raw, tactile, printed, imperfect, and handmade. Avoid a glossy modern finish."diff --git a/template/image-generation/futuristic-biophilic-atrium/1/cover.jpg b/template/image-generation/futuristic-biophilic-atrium/1/cover.jpgindex 11e5b70c..968822d3 100644Binary files a/template/image-generation/futuristic-biophilic-atrium/1/cover.jpg and b/template/image-generation/futuristic-biophilic-atrium/1/cover.jpg differdiff --git a/template/image-generation/graded-english-magazine-reading-page/1/cover.jpg b/template/image-generation/graded-english-magazine-reading-page/1/cover.jpgindex c472565f..26bcd055 100644Binary files a/template/image-generation/graded-english-magazine-reading-page/1/cover.jpg and b/template/image-generation/graded-english-magazine-reading-page/1/cover.jpg differdiff --git a/template/image-generation/grimm-fairy-tale-woodcut-poster/1/cover.jpg b/template/image-generation/grimm-fairy-tale-woodcut-poster/1/cover.jpgindex 0fd35f9e..e18d98f3 100644Binary files a/template/image-generation/grimm-fairy-tale-woodcut-poster/1/cover.jpg and b/template/image-generation/grimm-fairy-tale-woodcut-poster/1/cover.jpg differdiff --git a/template/image-generation/half-photo-half-crayon-memory-card/1/cover.jpg b/template/image-generation/half-photo-half-crayon-memory-card/1/cover.jpgindex ea0322de..13587845 100644Binary files a/template/image-generation/half-photo-half-crayon-memory-card/1/cover.jpg and b/template/image-generation/half-photo-half-crayon-memory-card/1/cover.jpg differdiff --git a/template/image-generation/half-photo-half-crayon-memory-card/1/template.json b/template/image-generation/half-photo-half-crayon-memory-card/1/template.jsonindex d96d2169..b4eaaae8 100644--- a/template/image-generation/half-photo-half-crayon-memory-card/1/template.json+++ b/template/image-generation/half-photo-half-crayon-memory-card/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Transform the uploaded photo into a vertical mixed-media memory card with a strict 50/50 split.\n\nKeep the original photo unchanged in the top half. In the bottom half, use textured off-white handmade paper and add a muted, irregular color patch matching the photo’s tones.\n\nRedraw the main subjects as a simple dark wax-crayon sketch with loose, imperfect lines and minimal details. Add a short handwritten English phrase and subtle Risograph grain.\n\nCreate a quiet, nostalgic Morandi-style aesthetic with generous negative space. Do not add extra elements or copy the reference composition exactly."+ "prompt": "将上传的照片转换为严格按 50/50 分割的竖版混合媒介回忆卡。\n\n上半部分保持原始照片不变。在下半部分,使用有纹理的灰白色手工纸,并添加与照片色调相匹配的柔和、不规则的色块。\n\n将主要人物或主体重新绘制为简单的深色蜡笔草图,具有松散、不完美的线条和最少的细节。添加简短的手写英文短语和细腻的 Risograph 孔版印刷颗粒。\n\n以宽敞的负空间营造出安静、怀旧的莫兰迪风格美学。不要添加额外的元素或完全复制参考构图。"},"en-US": {"prompt": "Transform the uploaded photo into a vertical mixed-media memory card with a strict 50/50 split.\n\nKeep the original photo unchanged in the top half. In the bottom half, use textured off-white handmade paper and add a muted, irregular color patch matching the photo’s tones.\n\nRedraw the main subjects as a simple dark wax-crayon sketch with loose, imperfect lines and minimal details. Add a short handwritten English phrase and subtle Risograph grain.\n\nCreate a quiet, nostalgic Morandi-style aesthetic with generous negative space. Do not add extra elements or copy the reference composition exactly."diff --git a/template/image-generation/hand-torn-paper-poster/1/cover.jpg b/template/image-generation/hand-torn-paper-poster/1/cover.jpgindex 40679a1e..cbdb2a1a 100644Binary files a/template/image-generation/hand-torn-paper-poster/1/cover.jpg and b/template/image-generation/hand-torn-paper-poster/1/cover.jpg differdiff --git a/template/image-generation/haute-couture-fashion-magazine-cover/1/cover.jpg b/template/image-generation/haute-couture-fashion-magazine-cover/1/cover.jpgindex 964b2f1f..22e7fb12 100644Binary files a/template/image-generation/haute-couture-fashion-magazine-cover/1/cover.jpg and b/template/image-generation/haute-couture-fashion-magazine-cover/1/cover.jpg differdiff --git a/template/image-generation/haute-couture-fashion-magazine-cover/1/template.json b/template/image-generation/haute-couture-fashion-magazine-cover/1/template.jsonindex 3af717a4..83cb96db 100644--- a/template/image-generation/haute-couture-fashion-magazine-cover/1/template.json+++ b/template/image-generation/haute-couture-fashion-magazine-cover/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Ultra high-fashion magazine cover, Louis Vuitton-style editorial. Close-up portrait of a confident woman with soft rose-gold hair and natural airy bangs, slightly wind-blown for movement. She is wearing a luxury summer outfit: a structured lightweight linen or silk jacket in warm golden-yellow tones, layered over a modest high-neck top, paired with a bold gold choker necklace and subtle statement earrings.\n\nFabric flows naturally with a summer breeze, slightly textured and breathable, capturing a premium seasonal feel. Styling is elegant, modest, and refined — no revealing clothing.\n\nLighting is high-end studio mixed with natural golden hour glow: warm highlights, soft shadows, luminous skin with glossy editorial finish.\n\nBackground is a rich summer gradient (sunset gold fading into soft coral or warm beige), clean but visually striking.\n\nComposition is dynamic and slightly cinematic: hair in motion, shallow depth of field, sharp focus on face.\n\nTypography: large elegant serif masthead \"Louis Vuitton\" at the top, bold cover line \"YOUR CHOICE ENDS HERE\" in premium editorial layout, minimal supporting text.\n\nUltra-realistic, hyper-detailed skin texture, 8K resolution, sharp focus, glossy magazine print quality, cinematic color grading, luxury fashion photography, no nudity, tasteful and editorial."+ "prompt": "超高时尚杂志封面,路易威登风格的社论。一位自信女性的特写肖像,有着柔软的玫瑰金色头发和自然通风的刘海,微风吹动,营造出动感。她穿着奢华的夏季服装:温暖的金黄色色调的结构轻质亚麻或丝绸夹克,搭配朴素的高领上衣,搭配大胆的金色项链和精致的个性耳环。\n\n面料随夏日微风自然飘逸,略带纹理且透气,捕捉优质季节感。造型优雅、端庄、精致——不穿暴露的衣服。\n\n高端工作室照明与自然的黄金时刻辉光混合在一起:温暖的高光、柔和的阴影、明亮的皮肤和光滑的编辑效果。\n\n背景是丰富的夏季渐变(夕阳金色逐渐变成柔软的珊瑚色或温暖的米色),干净但视觉上引人注目。\n\n构图充满活力,略带电影感:头发在移动,景深较浅,脸部焦点清晰。\n\n版式:顶部有大而优雅的衬线标头“Louis Vuitton”,粗体封面线“YOUR CHOICE ENDS HERE”采用高级编辑布局,最小化支持文本。\n\n超真实、超细致的皮肤纹理、8K 分辨率、锐利的焦点、光泽杂志印刷质量、电影调色、奢华时尚摄影、无裸露、品味和社论。"},"en-US": {"prompt": "Ultra high-fashion magazine cover, Louis Vuitton-style editorial. Close-up portrait of a confident woman with soft rose-gold hair and natural airy bangs, slightly wind-blown for movement. She is wearing a luxury summer outfit: a structured lightweight linen or silk jacket in warm golden-yellow tones, layered over a modest high-neck top, paired with a bold gold choker necklace and subtle statement earrings.\n\nFabric flows naturally with a summer breeze, slightly textured and breathable, capturing a premium seasonal feel. Styling is elegant, modest, and refined — no revealing clothing.\n\nLighting is high-end studio mixed with natural golden hour glow: warm highlights, soft shadows, luminous skin with glossy editorial finish.\n\nBackground is a rich summer gradient (sunset gold fading into soft coral or warm beige), clean but visually striking.\n\nComposition is dynamic and slightly cinematic: hair in motion, shallow depth of field, sharp focus on face.\n\nTypography: large elegant serif masthead \"Louis Vuitton\" at the top, bold cover line \"YOUR CHOICE ENDS HERE\" in premium editorial layout, minimal supporting text.\n\nUltra-realistic, hyper-detailed skin texture, 8K resolution, sharp focus, glossy magazine print quality, cinematic color grading, luxury fashion photography, no nudity, tasteful and editorial."diff --git a/template/image-generation/ink-double-exposure-portrait-poster/1/cover.jpg b/template/image-generation/ink-double-exposure-portrait-poster/1/cover.jpgindex 8903331e..0eab4bb6 100644Binary files a/template/image-generation/ink-double-exposure-portrait-poster/1/cover.jpg and b/template/image-generation/ink-double-exposure-portrait-poster/1/cover.jpg differdiff --git a/template/image-generation/ink-double-exposure-portrait-poster/1/template.json b/template/image-generation/ink-double-exposure-portrait-poster/1/template.jsonindex e4913d64..8d709afc 100644--- a/template/image-generation/ink-double-exposure-portrait-poster/1/template.json+++ b/template/image-generation/ink-double-exposure-portrait-poster/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "A cinematic character promotional poster of [SUBJECT], vertical composition (9:16), designed with a refined East-Asian ink aesthetic and high-end visual storytelling.\n\nSTRUCTURE:\nTop-heavy hierarchical layout. The upper half features a large, highly recognizable silhouette of [SUBJECT]'s head / face / mask / upper body, forming a bold, iconic primary shape. The silhouette should be instantly identifiable.\n\nThe middle-lower section contains the full-body version of [SUBJECT] as a secondary subject, standing in a stable pose or subtle action stance, forming the visual core.\n\nCOMPOSITION STYLE:\nInside the large silhouette and around the character, use double exposure and collage storytelling. Integrate multiple elements:\n- key scenes related to [SUBJECT]\n- symbolic imagery and environment\n- small narrative figures and interactions\n- supporting visual motifs\n\nBlend everything seamlessly using clouds, mist, ink diffusion, and negative space.\n\nVISUAL FLOW:\nCreate a continuous flowing visual path from top to bottom, connecting:\n- upper silhouette\n- inner collage elements\n- full-body subject\n\nEnsure smooth eye guidance and compositional cohesion.\n\nSIDE ELEMENTS:\nAdd balanced supporting elements on left and right sides to create tension, depth, and spatial variation.\n\nSTYLE & ATMOSPHERE:\n- Large areas of negative space\n- Ink-wash diffusion edges, soft fading, subtle fragmentation\n- Eastern aesthetic: balance of emptiness and detail\n- Calm, premium, restrained, cinematic tone\n\nQUALITY:\nUltra-detailed, high resolution, layered depth, soft lighting, atmospheric perspective, cohesive series-style design.\n\nOUTPUT:\n9:16 aspect ratio, poster-ready composition."+ "prompt": "《SUBJECT》电影角色宣传海报,垂直构图(9:16),采用精致的东亚水墨美学和高端视觉叙事设计。\n\n结构:\n头重脚轻的分层布局。上半部分以[SUBJECT]头部/面部/面具/上半身的大型、高度可识别的轮廓为特色,形成大胆、标志性的主要形状。轮廓应该可以立即识别。\n\n中下部分包含全身版的【SUBJECT】作为次要主体,以稳定的姿势或微妙的动作姿态站立,形成视觉核心。\n\n构图风格:\n在大型剪影内部和角色周围,使用双重曝光和拼贴讲故事。整合多个元素:\n- 与[SUBJECT]相关的关键场景\n- 象征意象和环境\n- 小叙事人物和互动\n- 支持视觉主题\n\n使用云、雾、墨水扩散和负空间无缝地融合一切。\n\n视觉流程:\n创建从上到下连续流动的视觉路径,连接:\n- 上部轮廓\n- 内部拼贴元素\n- 全身主题\n\n确保流畅的视线引导和构图凝聚力。\n\n侧面元素:\n在左右两侧添加平衡的支撑元素,以创造张力、深度和空间变化。\n\n风格与氛围:\n- 大面积的负空间\n- 水墨扩散边缘、柔和褪色、细微碎裂\n- 东方美学:空虚与细节的平衡\n- 平静、优质、内敛、电影般的基调\n\n质量:\n超细致、高分辨率、层次深度、柔和的灯光、大气的视角、连贯的系列式设计。\n\n输出:\n9:16 的宽高比,适合海报的构图。"},"en-US": {"prompt": "A cinematic character promotional poster of [SUBJECT], vertical composition (9:16), designed with a refined East-Asian ink aesthetic and high-end visual storytelling.\n\nSTRUCTURE:\nTop-heavy hierarchical layout. The upper half features a large, highly recognizable silhouette of [SUBJECT]'s head / face / mask / upper body, forming a bold, iconic primary shape. The silhouette should be instantly identifiable.\n\nThe middle-lower section contains the full-body version of [SUBJECT] as a secondary subject, standing in a stable pose or subtle action stance, forming the visual core.\n\nCOMPOSITION STYLE:\nInside the large silhouette and around the character, use double exposure and collage storytelling. Integrate multiple elements:\n- key scenes related to [SUBJECT]\n- symbolic imagery and environment\n- small narrative figures and interactions\n- supporting visual motifs\n\nBlend everything seamlessly using clouds, mist, ink diffusion, and negative space.\n\nVISUAL FLOW:\nCreate a continuous flowing visual path from top to bottom, connecting:\n- upper silhouette\n- inner collage elements\n- full-body subject\n\nEnsure smooth eye guidance and compositional cohesion.\n\nSIDE ELEMENTS:\nAdd balanced supporting elements on left and right sides to create tension, depth, and spatial variation.\n\nSTYLE & ATMOSPHERE:\n- Large areas of negative space\n- Ink-wash diffusion edges, soft fading, subtle fragmentation\n- Eastern aesthetic: balance of emptiness and detail\n- Calm, premium, restrained, cinematic tone\n\nQUALITY:\nUltra-detailed, high resolution, layered depth, soft lighting, atmospheric perspective, cohesive series-style design.\n\nOUTPUT:\n9:16 aspect ratio, poster-ready composition."diff --git a/template/image-generation/inside-box-farm-ad-poster/1/cover.jpg b/template/image-generation/inside-box-farm-ad-poster/1/cover.jpgindex d6631cce..67e41cbd 100644Binary files a/template/image-generation/inside-box-farm-ad-poster/1/cover.jpg and b/template/image-generation/inside-box-farm-ad-poster/1/cover.jpg differdiff --git a/template/image-generation/inside-cup-fisheye-summer-drink-ad/1/cover.jpg b/template/image-generation/inside-cup-fisheye-summer-drink-ad/1/cover.jpgindex 76a2f0c7..cbcf6eb9 100644Binary files a/template/image-generation/inside-cup-fisheye-summer-drink-ad/1/cover.jpg and b/template/image-generation/inside-cup-fisheye-summer-drink-ad/1/cover.jpg differdiff --git a/template/image-generation/isometric-miniature-kitchen/1/cover.jpg b/template/image-generation/isometric-miniature-kitchen/1/cover.jpgindex 1e42489a..d0c6462a 100644Binary files a/template/image-generation/isometric-miniature-kitchen/1/cover.jpg and b/template/image-generation/isometric-miniature-kitchen/1/cover.jpg differdiff --git a/template/image-generation/luxury-mechanical-watch-technical-guide/1/cover.jpg b/template/image-generation/luxury-mechanical-watch-technical-guide/1/cover.jpgindex 69b6be8a..b6cbc2ed 100644Binary files a/template/image-generation/luxury-mechanical-watch-technical-guide/1/cover.jpg and b/template/image-generation/luxury-mechanical-watch-technical-guide/1/cover.jpg differdiff --git a/template/image-generation/luxury-mechanical-watch-technical-guide/1/template.json b/template/image-generation/luxury-mechanical-watch-technical-guide/1/template.jsonindex 987245d3..dd84a856 100644--- a/template/image-generation/luxury-mechanical-watch-technical-guide/1/template.json+++ b/template/image-generation/luxury-mechanical-watch-technical-guide/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "2x2 grid 16:9, do this for 4 most expensive strangest watches ever made:\n\nclass Haute_Horlogerie_DNA:\n def __init__(self):\n self.subject = \"[TIMEPIECE]\"\n self.parents = {\n \"composition_parent\": \"Exploded movement diagram with transparent case\",\n \"material_parent\": \"Brushed titanium, sapphire crystal, rose gold gears, alligator leather\",\n \"graphic_parent\": \"Swiss manufacture technical brochure with elegant data panels\",\n \"atmosphere_parent\": \"Crisp daylight studio, pure white background, subtle reflection on polished surfaces\"\n }\n self.mutations = {\n \"semantic_mutation\": \"The balance wheel reveals a miniature cosmos ticking inside\",\n \"information_mutation\": \"Power reserve indicator, frequency, complication callouts, hand-finishing grades, assembly timeline\",\n \"medium_mutation\": \"Smooth matte premium paper with embossed logo\",\n \"scale_mutation\": \"Grain-level view of Côtes de Genève finishing and jewel bearings\"\n }\n self.style_mix = [0.30, 0.30, 0.25, 0.10, 0.05]\n\n def generate_subject(self):\n subject = \"\"\"\n [TIMEPIECE] shown in its full mechanical glory. The dial, hands, movement,\n and strap float in perfect alignment against a bright, clean background.\n Every gear and spring is highlighted with exacting clarity.\n \"\"\"\n return render(\n subject,\n format=\"luxury watch advertisement with technical insert\",\n title=\"[MODEL REFERENCE]\",\n subtitle=\"[MANUFACTURE / COLLECTION]\",\n constraints=\"bright white space, metallic brilliance, hyper-detailed, modern elegance\"\n )"+ "prompt": "生成一张 16:9 的 2×2 网格图,分别展示四款史上最昂贵、最奇特的高级机械腕表。\n\n每一格都以 [TIMEPIECE] 为主体,完整呈现表盘、指针、机芯和表带。各部件以爆炸分解图方式悬浮并精准对齐,透明表壳让内部结构清晰可见;每一枚齿轮、弹簧、摆轮和宝石轴承都要具有高精度细节。\n\n材质包括拉丝钛金属、蓝宝石水晶、玫瑰金齿轮和鳄鱼皮表带。摆轮内部呈现一个正在运转的微缩宇宙。加入动力储存、振频、复杂功能、手工修饰等级和装配时间线等技术标注,并展示日内瓦波纹与宝石轴承的微观细节。\n\n整体设计参考高级瑞士制表技术手册:纯白背景、清晰日光摄影棚照明、抛光表面带克制反射、优质哑光纸张质感和压印标志。每格配有优雅的数据面板,标题使用 [MODEL REFERENCE],副标题使用 [MANUFACTURE / COLLECTION]。\n\n风格比例参考:爆炸结构图 30%、高级材质表现 30%、技术手册信息设计 25%、明亮摄影棚氛围 10%、微观机械细节 5%。画面必须保持充足白色空间、金属光泽、超高细节和现代优雅感。"},"en-US": {"prompt": "2x2 grid 16:9, do this for 4 most expensive strangest watches ever made:\n\nclass Haute_Horlogerie_DNA:\n def __init__(self):\n self.subject = \"[TIMEPIECE]\"\n self.parents = {\n \"composition_parent\": \"Exploded movement diagram with transparent case\",\n \"material_parent\": \"Brushed titanium, sapphire crystal, rose gold gears, alligator leather\",\n \"graphic_parent\": \"Swiss manufacture technical brochure with elegant data panels\",\n \"atmosphere_parent\": \"Crisp daylight studio, pure white background, subtle reflection on polished surfaces\"\n }\n self.mutations = {\n \"semantic_mutation\": \"The balance wheel reveals a miniature cosmos ticking inside\",\n \"information_mutation\": \"Power reserve indicator, frequency, complication callouts, hand-finishing grades, assembly timeline\",\n \"medium_mutation\": \"Smooth matte premium paper with embossed logo\",\n \"scale_mutation\": \"Grain-level view of Côtes de Genève finishing and jewel bearings\"\n }\n self.style_mix = [0.30, 0.30, 0.25, 0.10, 0.05]\n\n def generate_subject(self):\n subject = \"\"\"\n [TIMEPIECE] shown in its full mechanical glory. The dial, hands, movement,\n and strap float in perfect alignment against a bright, clean background.\n Every gear and spring is highlighted with exacting clarity.\n \"\"\"\n return render(\n subject,\n format=\"luxury watch advertisement with technical insert\",\n title=\"[MODEL REFERENCE]\",\n subtitle=\"[MANUFACTURE / COLLECTION]\",\n constraints=\"bright white space, metallic brilliance, hyper-detailed, modern elegance\"\n )"diff --git a/template/image-generation/miniature-city-map-travel-poster/1/cover.jpg b/template/image-generation/miniature-city-map-travel-poster/1/cover.jpgindex b0250723..8de1a9a1 100644Binary files a/template/image-generation/miniature-city-map-travel-poster/1/cover.jpg and b/template/image-generation/miniature-city-map-travel-poster/1/cover.jpg differdiff --git a/template/image-generation/miniature-city-map-travel-poster/1/template.json b/template/image-generation/miniature-city-map-travel-poster/1/template.jsonindex 20fa7f5d..64916af4 100644--- a/template/image-generation/miniature-city-map-travel-poster/1/template.json+++ b/template/image-generation/miniature-city-map-travel-poster/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Create a highly detailed cinematic miniature tilt-shift travel scene of [CITY NAME] featuring a realistic [VEHICLE NAME] driving along a winding elevated road that emerges naturally from a printed vintage-style city map. The road should curve dramatically toward the background skyline and landmarks of [CITY NAME], while the vehicle remains the clear focal point in the foreground.\n\nBlend the real city seamlessly with the illustrated map surface so the road appears integrated into the map itself. Include recognizable local landmarks, waterways, architecture, vegetation, and atmosphere associated with [CITY NAME], but keep the composition clean and uncluttered.\n\nShow large bold typography of \"[CITY NAME]\" printed directly on the map in the foreground. Use warm golden-hour lighting, shallow depth of field, realistic textures, cinematic shadows, aerial perspective, and photorealistic detail. The overall aesthetic should feel like a premium Instagram travel poster mixed with a miniature diorama.\n\nAspect ratio 1:1."+ "prompt": "创建 [城市名称] 的高度详细的电影微型移轴旅行场景,其中包含逼真的 [车辆名称] 沿着蜿蜒的高架道路行驶,该道路从印刷的复古风格城市地图中自然出现。道路应向[城市名称]的背景天际线和地标急剧弯曲,而车辆仍然是前景中的清晰焦点。\n\n将真实城市与插图地图表面无缝融合,使道路看起来与地图本身融为一体。包括与[城市名称]相关的可识别的当地地标、水道、建筑、植被和氛围,但保持构图干净整洁。\n\n显示直接打印在前景地图上的大号粗体字体“[城市名称]”。使用温暖的黄金时段照明、浅景深、逼真的纹理、电影阴影、空中透视和逼真的细节。整体美感应该像是一张高级 Instagram 旅行海报与微型立体模型的混合体。\n\n长宽比 1:1。"},"en-US": {"prompt": "Create a highly detailed cinematic miniature tilt-shift travel scene of [CITY NAME] featuring a realistic [VEHICLE NAME] driving along a winding elevated road that emerges naturally from a printed vintage-style city map. The road should curve dramatically toward the background skyline and landmarks of [CITY NAME], while the vehicle remains the clear focal point in the foreground.\n\nBlend the real city seamlessly with the illustrated map surface so the road appears integrated into the map itself. Include recognizable local landmarks, waterways, architecture, vegetation, and atmosphere associated with [CITY NAME], but keep the composition clean and uncluttered.\n\nShow large bold typography of \"[CITY NAME]\" printed directly on the map in the foreground. Use warm golden-hour lighting, shallow depth of field, realistic textures, cinematic shadows, aerial perspective, and photorealistic detail. The overall aesthetic should feel like a premium Instagram travel poster mixed with a miniature diorama.\n\nAspect ratio 1:1."diff --git a/template/image-generation/minimal-abstract-geometric-deconstruction-poster/1/cover.jpg b/template/image-generation/minimal-abstract-geometric-deconstruction-poster/1/cover.jpgindex 8b0b6370..7d6a492c 100644Binary files a/template/image-generation/minimal-abstract-geometric-deconstruction-poster/1/cover.jpg and b/template/image-generation/minimal-abstract-geometric-deconstruction-poster/1/cover.jpg differdiff --git a/template/image-generation/minimal-conceptual-line-art-poster/1/cover.jpg b/template/image-generation/minimal-conceptual-line-art-poster/1/cover.jpgindex ca96d6aa..0f154f1f 100644Binary files a/template/image-generation/minimal-conceptual-line-art-poster/1/cover.jpg and b/template/image-generation/minimal-conceptual-line-art-poster/1/cover.jpg differdiff --git a/template/image-generation/modern-art-event-flyer/1/cover.jpg b/template/image-generation/modern-art-event-flyer/1/cover.jpgindex f15e7f2c..ef18d72f 100644Binary files a/template/image-generation/modern-art-event-flyer/1/cover.jpg and b/template/image-generation/modern-art-event-flyer/1/cover.jpg differdiff --git a/template/image-generation/paper-relief-neighborhood-grid/1/cover.jpg b/template/image-generation/paper-relief-neighborhood-grid/1/cover.jpgindex e0bf906d..d93075b9 100644Binary files a/template/image-generation/paper-relief-neighborhood-grid/1/cover.jpg and b/template/image-generation/paper-relief-neighborhood-grid/1/cover.jpg differdiff --git a/template/image-generation/playful-ceramic-ecommerce-hero/1/cover.jpg b/template/image-generation/playful-ceramic-ecommerce-hero/1/cover.jpgindex 7fcee1be..361b55b6 100644Binary files a/template/image-generation/playful-ceramic-ecommerce-hero/1/cover.jpg and b/template/image-generation/playful-ceramic-ecommerce-hero/1/cover.jpg differdiff --git a/template/image-generation/portrait-lighting-scheme-grid/1/cover.jpg b/template/image-generation/portrait-lighting-scheme-grid/1/cover.jpgindex e9ba675c..f124e62f 100644Binary files a/template/image-generation/portrait-lighting-scheme-grid/1/cover.jpg and b/template/image-generation/portrait-lighting-scheme-grid/1/cover.jpg differdiff --git a/template/image-generation/premium-food-commercial-storyboard/1/cover.jpg b/template/image-generation/premium-food-commercial-storyboard/1/cover.jpgindex 4283275a..32f13aa6 100644Binary files a/template/image-generation/premium-food-commercial-storyboard/1/cover.jpg and b/template/image-generation/premium-food-commercial-storyboard/1/cover.jpg differdiff --git a/template/image-generation/reflective-ski-gear-landing-page/1/cover.jpg b/template/image-generation/reflective-ski-gear-landing-page/1/cover.jpgindex ef829d0f..74b68460 100644Binary files a/template/image-generation/reflective-ski-gear-landing-page/1/cover.jpg and b/template/image-generation/reflective-ski-gear-landing-page/1/cover.jpg differdiff --git a/template/image-generation/rio-travel-ticket-paper-diorama-poster/1/cover.jpg b/template/image-generation/rio-travel-ticket-paper-diorama-poster/1/cover.jpgindex c92277ab..32d7b0f8 100644Binary files a/template/image-generation/rio-travel-ticket-paper-diorama-poster/1/cover.jpg and b/template/image-generation/rio-travel-ticket-paper-diorama-poster/1/cover.jpg differdiff --git a/template/image-generation/rio-travel-ticket-paper-diorama-poster/1/template.json b/template/image-generation/rio-travel-ticket-paper-diorama-poster/1/template.jsonindex c57c9c2f..473abf60 100644--- a/template/image-generation/rio-travel-ticket-paper-diorama-poster/1/template.json+++ b/template/image-generation/rio-travel-ticket-paper-diorama-poster/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Create a highly detailed, photorealistic miniature travel-poster diorama inspired by Rio de Janeiro, arranged as a handcrafted 3D paper scene on a warm ivory, slightly textured background.\n\nIn the foreground, a realistic human hand holds a vintage Brazilian travel ticket or Rio-themed transit card vertically on the left side. Give the card aged paper texture, subtle printing imperfections, elegant typography, and authentic-looking travel details. From behind the card, a miniature Rio de Janeiro landscape physically rises outward like an intricate pop-up diorama.\n\nMake Christ the Redeemer the dominant central landmark, positioned high above a miniature cityscape with lush green mountains surrounding it. Below, build a tiny realistic Rio street featuring a classic yellow taxi, colorful buildings, palm trees, pedestrians, cyclists, street lamps, tiled sidewalks, and small Brazilian urban details. Add Copacabana beach elements in the distance with tiny umbrellas, beachgoers, and a glimpse of the Atlantic Ocean. Layer the architecture and terrain so everything appears physically constructed from paper, wood, plaster, and miniature materials, with convincing depth, cast shadows, overlapping surfaces, and a slight three-quarter perspective.\n\nAround the main 3D scene, incorporate delicate black, charcoal, and muted sepia hand-drawn travel illustrations on the cream paper. Include a small Sugarloaf Mountain sketch in the upper left, an artistic Copacabana promenade illustration in the upper right, a detailed Selarón Steps sketch along the right side, and a small Ipanema beachfront skyline drawing near the bottom. Add subtle handwritten travel notes, tiny map markings, architectural outlines, compass symbols, postage-stamp details, and understated Brazilian travel annotations.\n\nKeep the composition refined rather than crowded. Blend realistic miniature photography with vintage travel-journal design, tactile paper fibers, faint ink bleed, imperfect hand-drawn lines, warm natural studio lighting, gentle shadows, subtle film grain, and a sophisticated cream, charcoal, muted green, ocean blue, and Brazilian yellow palette.\n\nThe final image should feel like a premium collectible Rio de Janeiro travel postcard transformed into a physical miniature world, with the central diorama sharply detailed and the surrounding illustrations slightly softer. Highly realistic human hand and fingers, believable miniature materials, cinematic product photography, editorial travel-magazine aesthetic, shallow depth of field, ultra-fine textures, photorealistic 3D details, vertical 4:5 composition, 8K quality."+ "prompt": "受里约热内卢启发,创建高度详细、逼真的微型旅行海报立体模型,在温暖的象牙色、略带纹理的背景上布置为手工制作的 3D 纸场景。\n\n在前景中,一只逼真的人手在左侧垂直握住一张老式巴西旅行票或里约主题交通卡。赋予卡片老化的纸张纹理、微妙的印刷瑕疵、优雅的排版和真实的旅行细节。从卡片后面,微型里约热内卢风景像一个复杂的弹出式立体模型一样向外升起。\n\n让救世主基督成为主要的中心地标,位于微型城市景观之上,周围环绕着郁郁葱葱的绿色山脉。下面,构建一条真实的里约小街,其中有一辆经典的黄色出租车、色彩缤纷的建筑、棕榈树、行人、骑自行车的人、路灯、瓷砖人行道和巴西小城市细节。远处添加了科帕卡巴纳海滩元素,包括小雨伞、海滩游客和大西洋的一瞥。将建筑和地形分层,使一切看起来都是由纸张、木材、石膏和微型材料物理构建的,具有令人信服的深度、投射阴影、重叠的表面和轻微的四分之三透视。\n\n围绕主要 3D 场景,在奶油色纸上融入精致的黑色、木炭色和柔和的棕褐色手绘旅行插图。包括左上角的小舒格洛夫山草图、右上角的科帕卡巴纳长廊艺术插图、右侧的详细塞拉隆阶梯草图以及靠近底部的伊帕内玛海滨天际线小图。添加微妙的手写旅行笔记、微小的地图标记、建筑轮廓、指南针符号、邮票细节和低调的巴西旅行注释。\n\n保持构图精致而不是拥挤。将逼真的微型摄影与复古旅行日记设计、触感纸纤维、微弱的墨水渗色、不完美的手绘线条、温暖的自然工作室灯光、柔和的阴影、微妙的胶片颗粒以及精致的奶油色、木炭色、柔和的绿色、海洋蓝色和巴西黄色调色板融为一体。\n\n最终图像应该感觉像是一张高级收藏级里约热内卢旅行明信片转变成一个物理微型世界,中央立体模型清晰细致,周围的插图稍微柔和。高度逼真的人手和手指、可信的微缩材质、电影产品摄影、编辑旅游杂志美学、浅景深、超精细纹理、逼真的 3D 细节、垂直 4:5 构图、8K 质量。"},"en-US": {"prompt": "Create a highly detailed, photorealistic miniature travel-poster diorama inspired by Rio de Janeiro, arranged as a handcrafted 3D paper scene on a warm ivory, slightly textured background.\n\nIn the foreground, a realistic human hand holds a vintage Brazilian travel ticket or Rio-themed transit card vertically on the left side. Give the card aged paper texture, subtle printing imperfections, elegant typography, and authentic-looking travel details. From behind the card, a miniature Rio de Janeiro landscape physically rises outward like an intricate pop-up diorama.\n\nMake Christ the Redeemer the dominant central landmark, positioned high above a miniature cityscape with lush green mountains surrounding it. Below, build a tiny realistic Rio street featuring a classic yellow taxi, colorful buildings, palm trees, pedestrians, cyclists, street lamps, tiled sidewalks, and small Brazilian urban details. Add Copacabana beach elements in the distance with tiny umbrellas, beachgoers, and a glimpse of the Atlantic Ocean. Layer the architecture and terrain so everything appears physically constructed from paper, wood, plaster, and miniature materials, with convincing depth, cast shadows, overlapping surfaces, and a slight three-quarter perspective.\n\nAround the main 3D scene, incorporate delicate black, charcoal, and muted sepia hand-drawn travel illustrations on the cream paper. Include a small Sugarloaf Mountain sketch in the upper left, an artistic Copacabana promenade illustration in the upper right, a detailed Selarón Steps sketch along the right side, and a small Ipanema beachfront skyline drawing near the bottom. Add subtle handwritten travel notes, tiny map markings, architectural outlines, compass symbols, postage-stamp details, and understated Brazilian travel annotations.\n\nKeep the composition refined rather than crowded. Blend realistic miniature photography with vintage travel-journal design, tactile paper fibers, faint ink bleed, imperfect hand-drawn lines, warm natural studio lighting, gentle shadows, subtle film grain, and a sophisticated cream, charcoal, muted green, ocean blue, and Brazilian yellow palette.\n\nThe final image should feel like a premium collectible Rio de Janeiro travel postcard transformed into a physical miniature world, with the central diorama sharply detailed and the surrounding illustrations slightly softer. Highly realistic human hand and fingers, believable miniature materials, cinematic product photography, editorial travel-magazine aesthetic, shallow depth of field, ultra-fine textures, photorealistic 3D details, vertical 4:5 composition, 8K quality."diff --git a/template/image-generation/rubber-stamp-travel-field-notes/1/cover.jpg b/template/image-generation/rubber-stamp-travel-field-notes/1/cover.jpgindex aca4f9f3..1f583fb7 100644Binary files a/template/image-generation/rubber-stamp-travel-field-notes/1/cover.jpg and b/template/image-generation/rubber-stamp-travel-field-notes/1/cover.jpg differdiff --git a/template/image-generation/rubber-stamp-travel-field-notes/1/template.json b/template/image-generation/rubber-stamp-travel-field-notes/1/template.jsonindex 0070c26c..f615763c 100644--- a/template/image-generation/rubber-stamp-travel-field-notes/1/template.json+++ b/template/image-generation/rubber-stamp-travel-field-notes/1/template.json@@ -34,7 +34,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Create a separate 4:3 landscape Rubber Stamp Travel Field Notes Poster for each uploaded photo. Never combine photos. Blend the authentic original photograph on the left (about 58%) into a warm aged-paper field-note area on the right (about 42%) without an obvious divider.\n\nPreserve the photograph's recognizable subject, perspective, terrain, architecture, people, spatial relationships, natural light, shadows, textures, and atmosphere. Allow only natural cropping and subtle editorial balancing, softened highlights, restrained color, and very fine film grain. Do not stretch, move, replace, redesign, or redraw the main subject; avoid HDR and artificial clarity.\n\nThe right side is believable matte off-white notebook paper with subtle fibers, fine grain, faint handling marks, and generous unprinted space. In its lower-middle area, place a compact imperfect multi-color rubber-stamp impression occupying about 30–38% of the paper height. Study the photo and reduce only its most recognizable architecture, silhouette, terrain, shoreline or road direction, vegetation, settlement layers, and one or two key foreground shapes. Remove crowds, cars, repetitive windows, dense buildings, tiny vegetation, and decorative clutter. It must look physically carved and stamped, not like a miniature digital illustration.\n\nExtract 2–4 restrained spot colors from the photo, such as carbon black, deep green, muted brick, ochre, slate blue, taupe, or other naturally appropriate colors. Render separate physical ink layers with slightly uneven pressure, tiny gaps, dry areas, paper showing through, rough carved edges, irregular line thickness, contour breaks, granular ink, faint ghosting, and slight natural misregistration. Avoid smooth vectors, gradients, perfect alignment, or exaggerated distress.\n\nAdd only quiet field-note text based on the actual location: English location name, a number, three short English keywords, and Gregorian year. Use small understated typewriter or archival typography with slight mechanical irregularity. Spell all text correctly; no slogans, brands, tourist phrases, decorative quotes, or excess copy.\n\nOverall feel: a real travel photograph mounted beside a handmade field stamp in an architect's notebook; tactile, documentary, restrained, collectible, and naturally imperfect. Avoid dividing lines, circular seals, postage borders, wax seals, stickers, souvenir-card templates, cartoons, 3D rendering, plastic texture, glossy gradients, excess saturation, dense text, decorative clutter, and alteration of the source photo."+ "prompt": "为每张上传的照片创建单独的 4:3 横向橡皮图章旅行笔记海报。切勿合并照片。将左侧真实的原始照片(约 58%)融合到右侧温暖的旧纸田野笔记区域(约 42%)中,没有明显的分隔线。\n\n保留照片可识别的主题、视角、地形、建筑、人物、空间关系、自然光、阴影、纹理和氛围。只允许自然裁剪和微妙的编辑平衡、柔和的高光、克制的色彩和非常精细的胶片颗粒。请勿拉伸、移动、替换、重新设计或重新绘制主要主题;避免 HDR 和人工清晰度。\n\n右侧是可信的哑光灰白色笔记本纸,具有微妙的纤维、细小的纹理、微弱的处理痕迹和大量的未印刷空间。在其中下区域,放置一个紧凑的、不完美的多色橡皮图章印记,占据纸张高度的 30-38% 左右。研究照片并仅减少其最容易识别的建筑、轮廓、地形、海岸线或道路方向、植被、聚落层以及一两个关键前景形状。清除人群、汽车、重复的窗户、密集的建筑物、微小的植被和杂乱的装饰。它必须看起来是实体雕刻和压印的,而不是像微型数字插图。\n\n从照片中提取 2-4 种受限制的专色,例如炭黑、深绿色、柔和的砖色、赭色、石板蓝、灰褐色或其他自然合适的颜色。渲染单独的物理墨水层,压力略有不均匀、微小间隙、干燥区域、纸张透过、粗糙的雕刻边缘、不规则的线条粗细、轮廓断裂、颗粒状墨水、微弱的重影和轻微的自然重影。避免平滑矢量、渐变、完美对齐或夸张的痛苦。\n\n根据实际位置仅添加安静的字段注释文本:英文位置名称、数字、三个简短的英文关键字和公历年份。使用小型低调打字机或带有轻微机械不规则性的档案排版。正确拼写所有文本;没有口号、品牌、旅游短语、装饰性引言或多余的文案。\n\n整体感觉:一张真实的旅行照片,贴在建筑师笔记本上手工制作的野外印章旁边;触觉的、记录的、内敛的、具有收藏价值的,而且天生不完美。避免分割线、圆形密封、邮资边框、蜡封、贴纸、纪念卡模板、卡通、3D 渲染、塑料纹理、光泽渐变、过度饱和度、密集文本、装饰杂乱以及对源照片的更改。"},"en-US": {"prompt": "Create a separate 4:3 landscape Rubber Stamp Travel Field Notes Poster for each uploaded photo. Never combine photos. Blend the authentic original photograph on the left (about 58%) into a warm aged-paper field-note area on the right (about 42%) without an obvious divider.\n\nPreserve the photograph's recognizable subject, perspective, terrain, architecture, people, spatial relationships, natural light, shadows, textures, and atmosphere. Allow only natural cropping and subtle editorial balancing, softened highlights, restrained color, and very fine film grain. Do not stretch, move, replace, redesign, or redraw the main subject; avoid HDR and artificial clarity.\n\nThe right side is believable matte off-white notebook paper with subtle fibers, fine grain, faint handling marks, and generous unprinted space. In its lower-middle area, place a compact imperfect multi-color rubber-stamp impression occupying about 30–38% of the paper height. Study the photo and reduce only its most recognizable architecture, silhouette, terrain, shoreline or road direction, vegetation, settlement layers, and one or two key foreground shapes. Remove crowds, cars, repetitive windows, dense buildings, tiny vegetation, and decorative clutter. It must look physically carved and stamped, not like a miniature digital illustration.\n\nExtract 2–4 restrained spot colors from the photo, such as carbon black, deep green, muted brick, ochre, slate blue, taupe, or other naturally appropriate colors. Render separate physical ink layers with slightly uneven pressure, tiny gaps, dry areas, paper showing through, rough carved edges, irregular line thickness, contour breaks, granular ink, faint ghosting, and slight natural misregistration. Avoid smooth vectors, gradients, perfect alignment, or exaggerated distress.\n\nAdd only quiet field-note text based on the actual location: English location name, a number, three short English keywords, and Gregorian year. Use small understated typewriter or archival typography with slight mechanical irregularity. Spell all text correctly; no slogans, brands, tourist phrases, decorative quotes, or excess copy.\n\nOverall feel: a real travel photograph mounted beside a handmade field stamp in an architect's notebook; tactile, documentary, restrained, collectible, and naturally imperfect. Avoid dividing lines, circular seals, postage borders, wax seals, stickers, souvenir-card templates, cartoons, 3D rendering, plastic texture, glossy gradients, excess saturation, dense text, decorative clutter, and alteration of the source photo."diff --git a/template/image-generation/same-face-twelve-hairstyles/1/cover.jpg b/template/image-generation/same-face-twelve-hairstyles/1/cover.jpgindex 0ddf235c..2065c853 100644Binary files a/template/image-generation/same-face-twelve-hairstyles/1/cover.jpg and b/template/image-generation/same-face-twelve-hairstyles/1/cover.jpg differdiff --git a/template/image-generation/same-face-twelve-hairstyles/1/template.json b/template/image-generation/same-face-twelve-hairstyles/1/template.jsonindex 40f32596..45c4e5f2 100644--- a/template/image-generation/same-face-twelve-hairstyles/1/template.json+++ b/template/image-generation/same-face-twelve-hairstyles/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Create a 12-panel grid (3 columns × 4 rows, numbered 1 to 12) showing the SAME person from the reference photo with 12 different hairstyles. This is a hairstyle lookbook. Final image aspect ratio: 4:5 (vertical/portrait).\nTHE ONLY THING THAT CHANGES BETWEEN PANELS IS THE HAIR ON THE HEAD (shape, style and length only). Everything else stays exactly as in the reference photo.\nIdentity Anchor (Critical)\nThe face must be IDENTICAL to the reference photo in every single panel. Preserve exactly: facial bone structure, jawline, cheekbones, nose shape, lips, eye shape and spacing, eyebrows, skin tone, skin texture (pores, natural imperfections), and overall facial proportions. This is the same real person in all 12 frames. Do NOT beautify, slim, or alter the face. Same age, same expression as in the reference.\nMandatory Rules (Do Not Violate)\n- NO SUNGLASSES. Eyes must be fully visible in all 12 panels.\n- NO ENVIRONMENTAL BACKGROUNDS. Every panel must have a plain, uniform, solid light grey studio backdrop with zero objects, zero textures, zero gradients. Just flat neutral grey.\n- Hair COLOR stays exactly as it appears in the reference photo in all 12 panels. Only the shape, length and style changes, never the color.\nKeep Identical in Every Panel (Do Not Change)\n- MAKEUP AND SKIN: If the person in the reference photo wears makeup, replicate it identically in every panel. Same lip color, same eye makeup, same brow grooming. If they wear no makeup, keep all panels makeup-free. Do NOT add, remove, or alter makeup between panels.\n- Clothing: the same clothing visible in the reference photo, replicated exactly.\n- Accessories: preserve ALL visible accessories from the reference photo (earrings, necklaces, rings, bracelets, piercings, watch, glasses, etc.). Do not omit, resize, recolor, or restyle any accessory. If the person wears prescription glasses (not sunglasses), keep them in every panel.\n- Background: plain solid light grey studio backdrop in every panel. No room, no furniture, no environment.\nThe 12 Hairstyles\n1. Pixie cut: very short, textured, slightly tousled on top with tapered sides and nape\n2. Classic bob: chin-length, straight, blunt ends, clean middle part\n3. Long layered waves: past the shoulders, soft voluminous waves with face-framing layers\n4. Sleek low bun: hair pulled back smoothly into a tight low bun at the nape, no flyaways\n5. Curtain bangs with medium-length hair: soft parted fringe framing the face, hair falling just past the shoulders\n6. High ponytail: hair pulled up into a sleek high ponytail, smooth crown, length falling behind\n7. French bob: short bob ending at the jawline with a soft blunt micro-fringe across the forehead\n8. Long straight hair with middle part: very long, sleek, pin-straight, falling well past the shoulders\n9. Shaggy wolf cut: medium length, heavy layers, choppy fringe, textured and voluminous with a slightly wild look\n10. Elegant updo: hair swept up into a polished chignon with soft face-framing tendrils\n11. Short curly crop: short voluminous curls all over, natural texture, tapered at the sides\n12. Side-swept Hollywood waves: long glamorous deep side part, sculpted vintage waves cascading over one shoulder\nPhotographic Specs\nShot on a Canon EOS R5 with an 85mm f/1.4 lens, studio portrait lighting (soft key light, subtle fill), shallow depth of field with sharp focus on the face. PLAIN SOLID LIGHT GREY STUDIO BACKGROUND in every panel. Natural skin rendering with visible pores and realistic hair strands (no plastic or CGI look). Consistent lighting, color grading and exposure across all 12 panels. Photorealistic, high detail, hyperrealistic, 8K. No illustration, no painterly effect, no over-smoothing. NO SUNGLASSES.\nEach panel clearly numbered 1 to 12 in the top-left corner. Overall output aspect ratio 4:5."+ "prompt": "创建一个 12 格网格(3 列 × 4 行,编号 1 到 12),显示参考照片中的同一个人具有 12 种不同的发型。这是一本发型手册。最终图像长宽比:4:5(垂直/纵向)。\n面板之间唯一改变的是头上的头发(仅形状、样式和长度)。其他一切都与参考照片中的完全相同。\n身份锚(关键)\n脸部必须与每个面板中的参考照片相同。准确保留:面部骨骼结构、下颌线、颧骨、鼻子形状、嘴唇、眼睛形状和间距、眉毛、肤色、皮肤纹理(毛孔、自然瑕疵)以及整体面部比例。这 12 帧中都是同一个人。请勿美化、瘦身或改变脸部。与参考照片中的年龄相同,表情相同。\n强制性规则(请勿违反)\n- 禁止佩戴太阳镜。眼睛在所有 12 个面板中必须完全可见。\n- 无环境背景。每个面板必须有一个朴素、统一、纯浅灰色工作室背景,零对象、零纹理、零渐变。只是平坦的中性灰色。\n- 头发颜色在所有 12 个面板中均与参考照片中显示的颜色完全相同。只是形状、长度和款式发生变化,颜色却没有变化。\n每个面板保持相同(不要更改)\n- 化妆和皮肤:如果参考照片中的人化妆,请在每个面板中进行相同的复制。一样的唇色,一样的眼妆,一样的眉毛。如果他们没有化妆,请保持所有面板不化妆。请勿添加、移除或更改面板之间的装饰。\n- 服装:与参考照片中可见的相同服装,完全复制。\n- 配饰:保留参考照片中的所有可见配饰(耳环、项链、戒指、手链、穿孔、手表、眼镜等)。请勿遗漏、调整任何配件的大小、重新着色或重新设计任何配件。如果此人佩戴处方眼镜(不是太阳镜),请将其放在每个面板中。\n- 背景:每个面板中都有纯色浅灰色工作室背景。没有房间,没有家具,没有环境。\n12种发型\n1.精灵剪裁:非常短,有纹理,顶部略微凌乱,两侧和颈背呈锥形\n2. 经典鲍勃:长度到下巴,直,末端钝,中间部分干净\n3. 长层次波浪:过肩、柔软、丰盈的波浪,具有修饰脸型的层次感\n4. 光滑的低发髻:头发在后颈处平滑地向后拉成一个紧致的低发髻,没有飞扬的现象\n5. 中长发的帘式刘海:柔软的分叉刘海勾勒出脸部轮廓,头发刚刚过肩\n6.高马尾辫:将头发挽成光滑的高马尾辫,头顶光滑,长度落在后面\n7. 法式鲍勃:短鲍勃在下巴处结束,前额有柔软的钝微刘海\n8. 中分长直发:非常长,光滑,直,垂过肩膀\n9. 蓬松狼尾剪:中等长度,厚重的层次,凌乱碎刘海,质感丰盈,略显野性\n10. 优雅的高髻:将头发盘成光滑的发髻,带有柔软的脸型卷须\n11. 短卷发:满头短卷发,自然纹理,两侧逐渐变细\n12.好莱坞侧分波浪卷:迷人的深侧长发,雕刻般的复古波浪层叠在肩上\n摄影规格\n使用佳能 EOS R5 拍摄,配备 85mm f/1.4 镜头,工作室人像照明(柔和的主光,微妙的填充),浅景深,脸部锐利对焦。每个面板上都有纯色浅灰色工作室背景。自然的皮肤渲染,具有可见的毛孔和逼真的发丝(无塑料或 CGI 外观)。所有 12 个面板的照明、颜色分级和曝光一致。逼真、高细节、超写实、8K。没有插图,没有绘画效果,没有过度平滑。没有太阳镜。\n每个面板的左上角都清楚地标有 1 到 12 的编号。整体输出长宽比4:5。"},"en-US": {"prompt": "Create a 12-panel grid (3 columns × 4 rows, numbered 1 to 12) showing the SAME person from the reference photo with 12 different hairstyles. This is a hairstyle lookbook. Final image aspect ratio: 4:5 (vertical/portrait).\nTHE ONLY THING THAT CHANGES BETWEEN PANELS IS THE HAIR ON THE HEAD (shape, style and length only). Everything else stays exactly as in the reference photo.\nIdentity Anchor (Critical)\nThe face must be IDENTICAL to the reference photo in every single panel. Preserve exactly: facial bone structure, jawline, cheekbones, nose shape, lips, eye shape and spacing, eyebrows, skin tone, skin texture (pores, natural imperfections), and overall facial proportions. This is the same real person in all 12 frames. Do NOT beautify, slim, or alter the face. Same age, same expression as in the reference.\nMandatory Rules (Do Not Violate)\n- NO SUNGLASSES. Eyes must be fully visible in all 12 panels.\n- NO ENVIRONMENTAL BACKGROUNDS. Every panel must have a plain, uniform, solid light grey studio backdrop with zero objects, zero textures, zero gradients. Just flat neutral grey.\n- Hair COLOR stays exactly as it appears in the reference photo in all 12 panels. Only the shape, length and style changes, never the color.\nKeep Identical in Every Panel (Do Not Change)\n- MAKEUP AND SKIN: If the person in the reference photo wears makeup, replicate it identically in every panel. Same lip color, same eye makeup, same brow grooming. If they wear no makeup, keep all panels makeup-free. Do NOT add, remove, or alter makeup between panels.\n- Clothing: the same clothing visible in the reference photo, replicated exactly.\n- Accessories: preserve ALL visible accessories from the reference photo (earrings, necklaces, rings, bracelets, piercings, watch, glasses, etc.). Do not omit, resize, recolor, or restyle any accessory. If the person wears prescription glasses (not sunglasses), keep them in every panel.\n- Background: plain solid light grey studio backdrop in every panel. No room, no furniture, no environment.\nThe 12 Hairstyles\n1. Pixie cut: very short, textured, slightly tousled on top with tapered sides and nape\n2. Classic bob: chin-length, straight, blunt ends, clean middle part\n3. Long layered waves: past the shoulders, soft voluminous waves with face-framing layers\n4. Sleek low bun: hair pulled back smoothly into a tight low bun at the nape, no flyaways\n5. Curtain bangs with medium-length hair: soft parted fringe framing the face, hair falling just past the shoulders\n6. High ponytail: hair pulled up into a sleek high ponytail, smooth crown, length falling behind\n7. French bob: short bob ending at the jawline with a soft blunt micro-fringe across the forehead\n8. Long straight hair with middle part: very long, sleek, pin-straight, falling well past the shoulders\n9. Shaggy wolf cut: medium length, heavy layers, choppy fringe, textured and voluminous with a slightly wild look\n10. Elegant updo: hair swept up into a polished chignon with soft face-framing tendrils\n11. Short curly crop: short voluminous curls all over, natural texture, tapered at the sides\n12. Side-swept Hollywood waves: long glamorous deep side part, sculpted vintage waves cascading over one shoulder\nPhotographic Specs\nShot on a Canon EOS R5 with an 85mm f/1.4 lens, studio portrait lighting (soft key light, subtle fill), shallow depth of field with sharp focus on the face. PLAIN SOLID LIGHT GREY STUDIO BACKGROUND in every panel. Natural skin rendering with visible pores and realistic hair strands (no plastic or CGI look). Consistent lighting, color grading and exposure across all 12 panels. Photorealistic, high detail, hyperrealistic, 8K. No illustration, no painterly effect, no over-smoothing. NO SUNGLASSES.\nEach panel clearly numbered 1 to 12 in the top-left corner. Overall output aspect ratio 4:5."diff --git a/template/image-generation/six-panel-miniature-beverage-ad/1/cover.jpg b/template/image-generation/six-panel-miniature-beverage-ad/1/cover.jpgindex e42c5398..812465c2 100644Binary files a/template/image-generation/six-panel-miniature-beverage-ad/1/cover.jpg and b/template/image-generation/six-panel-miniature-beverage-ad/1/cover.jpg differdiff --git a/template/image-generation/six-panel-miniature-beverage-ad/1/template.json b/template/image-generation/six-panel-miniature-beverage-ad/1/template.jsonindex dd450354..d69cfbcc 100644--- a/template/image-generation/six-panel-miniature-beverage-ad/1/template.json+++ b/template/image-generation/six-panel-miniature-beverage-ad/1/template.json@@ -34,7 +34,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Create a Cannes-level premium summer beverage campaign poster for a fictional lemon drink brand called \"LIMORA\". Use a strict 2-column by 3-row grid with six equal panels, thin white dividers, a consistent horizon, beach-and-ocean setting, and unified bright upper-left seaside lighting. Across all panels, show the same ultra-realistic miniature young woman in a fitted green floral summer dress and white sandals interacting with oversized lemons, juice, sorbet, and the final drink. Keep her face, hair, anatomy, clothing, and scale consistent. The lemon beverage system is the visual hero.\n\nBuild one continuous visual rhythm from panel 1 to panel 6: touch, recline, squeeze, travel, embrace, taste. Repeat directional cues through hair, limbs, droplets, spoon angle, lemon slices, straw, and giant-hand placement.\n\nPanel 1: she hugs a giant whole lemon while a realistic adult hand lowers it from above.\nPanel 2: she reclines inside a halved lemon like a luxury beach chaise, with sunglasses and a tiny parasol pick; a lemon slice descends like a sun.\nPanel 3: a giant hand squeezes a cut lemon, sending translucent juice streams and droplets around her dynamic reaction.\nPanel 4: she rides a grounded refined wooden cart piled with lemons, pulled by a premium lemon-shaped creature or rolling harness.\nPanel 5: the hero panel. A tall LIMORA lemonade glass dominates, with clear glass, ice, lemon slices, pale sparkling liquid, condensation, green straw, and cocktail umbrella. She hugs the glass joyfully.\nPanel 6: she sits in a halved lemon while a polished spoon lowers glossy lemon sorbet or crushed ice.\n\nRender hyper-real citrus peel pores, wet translucent pulp, physically plausible juice, ice refraction, glass condensation, sand pressure marks, natural skin texture, correct hands and fingers, realistic fabric and hair. Use lemon yellow as the dominant color with citrus green, turquoise sea, pale blue sky, beach beige, and crisp white. Keep original minimal branding only on the hero glass, optionally adding \"LIMORA — Bright in Motion\" below the grid.\n\nStyle: luxury FMCG campaign, surreal miniature advertising photography, editorial storyboard energy, elegant humor, polished magazine finish, strong product realism. Avoid copied text, real brands, cartoon or toy styling, distorted anatomy, extra or missing fingers, plastic skin, fake citrus, cloudy glass, inconsistent identity, lighting or horizon, messy grid, clutter, meme styling, low resolution, oversaturation, watermarks, and distorted giant-hand perspective."+ "prompt": "为名为“LIMORA”的虚构柠檬饮料品牌制作戛纳级别的优质夏季饮料宣传海报。使用严格的 2 列 x 3 行网格,其中有六个相等的面板、薄的白色分隔线、一致的地平线、海滩和海洋环境以及统一的明亮的左上海滨照明。在所有面板上,都展示了同样的超现实微型年轻女子,穿着合身的绿色花卉夏装和白色凉鞋,与超大的柠檬、果汁、果汁冰糕和最后的饮料相互作用。保持她的脸部、头发、身体结构、服装和比例一致。柠檬饮料产品始终是视觉主角。\n\n从面板 1 到面板 6 构建一种连续的视觉节奏:触摸、倚靠、挤压、移动、拥抱、品尝。通过头发、四肢、水滴、勺子角度、柠檬片、吸管和巨手的位置重复方向提示。\n\n第 1 幅:她拥抱一个巨大的整个柠檬,而一只写实的成人手将其从上方放下。\n第 2 幅:她像豪华沙滩躺椅一样斜倚在半个柠檬里,戴着太阳镜和一把小遮阳伞;一片柠檬片像太阳一样落下。\n第 3 幅:一只巨大的手挤压切好的柠檬,在她的动态反应周围喷出半透明的果汁流和水滴。\n第 4 幅:她骑着一辆装满柠檬的精致木车,由精致的柠檬造型生物或滚动拖具拉动。\n面板 5:英雄面板。一个高高的 LIMORA 柠檬水杯占主导地位,里面有透明玻璃、冰块、柠檬片、浅色起泡液体、冷凝水、绿色吸管和鸡尾酒伞。她高兴地拥抱着玻璃杯。\n第 6 幅:她坐在半个柠檬中,同时用抛光的勺子降低光滑的柠檬冰糕或碎冰。\n\n渲染超真实的柑橘皮毛孔、湿润的半透明果肉、物理上合理的果汁、冰折射、玻璃凝结、沙压痕、自然的皮肤纹理、正确的手和手指、逼真的织物和头发。以柠檬黄为主色调,搭配柑橘绿、碧蓝海水、淡蓝天空、海滩米色和清爽的白色。只在主视觉玻璃杯上保留最少量的原创品牌元素,可以选择在网格下方添加“LIMORA — Bright in Motion”。\n\n风格:奢华的快消品宣传活动、超现实的微型广告摄影、编辑故事板的能量、优雅的幽默、精美的杂志完成、强烈的产品现实主义。避免复制文本、真实品牌、卡通或玩具造型、扭曲的解剖结构、多余或缺失的手指、塑料皮肤、假柑橘、浑浊的玻璃、不一致的身份、照明或地平线、凌乱的网格、混乱、模因造型、低分辨率、过饱和、水印和扭曲的巨手视角。"},"en-US": {"prompt": "Create a Cannes-level premium summer beverage campaign poster for a fictional lemon drink brand called \"LIMORA\". Use a strict 2-column by 3-row grid with six equal panels, thin white dividers, a consistent horizon, beach-and-ocean setting, and unified bright upper-left seaside lighting. Across all panels, show the same ultra-realistic miniature young woman in a fitted green floral summer dress and white sandals interacting with oversized lemons, juice, sorbet, and the final drink. Keep her face, hair, anatomy, clothing, and scale consistent. The lemon beverage system is the visual hero.\n\nBuild one continuous visual rhythm from panel 1 to panel 6: touch, recline, squeeze, travel, embrace, taste. Repeat directional cues through hair, limbs, droplets, spoon angle, lemon slices, straw, and giant-hand placement.\n\nPanel 1: she hugs a giant whole lemon while a realistic adult hand lowers it from above.\nPanel 2: she reclines inside a halved lemon like a luxury beach chaise, with sunglasses and a tiny parasol pick; a lemon slice descends like a sun.\nPanel 3: a giant hand squeezes a cut lemon, sending translucent juice streams and droplets around her dynamic reaction.\nPanel 4: she rides a grounded refined wooden cart piled with lemons, pulled by a premium lemon-shaped creature or rolling harness.\nPanel 5: the hero panel. A tall LIMORA lemonade glass dominates, with clear glass, ice, lemon slices, pale sparkling liquid, condensation, green straw, and cocktail umbrella. She hugs the glass joyfully.\nPanel 6: she sits in a halved lemon while a polished spoon lowers glossy lemon sorbet or crushed ice.\n\nRender hyper-real citrus peel pores, wet translucent pulp, physically plausible juice, ice refraction, glass condensation, sand pressure marks, natural skin texture, correct hands and fingers, realistic fabric and hair. Use lemon yellow as the dominant color with citrus green, turquoise sea, pale blue sky, beach beige, and crisp white. Keep original minimal branding only on the hero glass, optionally adding \"LIMORA — Bright in Motion\" below the grid.\n\nStyle: luxury FMCG campaign, surreal miniature advertising photography, editorial storyboard energy, elegant humor, polished magazine finish, strong product realism. Avoid copied text, real brands, cartoon or toy styling, distorted anatomy, extra or missing fingers, plastic skin, fake citrus, cloudy glass, inconsistent identity, lighting or horizon, messy grid, clutter, meme styling, low resolution, oversaturation, watermarks, and distorted giant-hand perspective."diff --git a/template/image-generation/social-media-giveaway-banner/1/cover.jpg b/template/image-generation/social-media-giveaway-banner/1/cover.jpgindex 3b2ea5ae..3aee937b 100644Binary files a/template/image-generation/social-media-giveaway-banner/1/cover.jpg and b/template/image-generation/social-media-giveaway-banner/1/cover.jpg differdiff --git a/template/image-generation/solar-terms-editorial-visual/1/cover.jpg b/template/image-generation/solar-terms-editorial-visual/1/cover.jpgindex ba2eacc6..33c38b1e 100644Binary files a/template/image-generation/solar-terms-editorial-visual/1/cover.jpg and b/template/image-generation/solar-terms-editorial-visual/1/cover.jpg differdiff --git a/template/image-generation/split-canvas-editorial-art-poster/1/cover.jpg b/template/image-generation/split-canvas-editorial-art-poster/1/cover.jpgindex 647e00fa..0ab6179a 100644Binary files a/template/image-generation/split-canvas-editorial-art-poster/1/cover.jpg and b/template/image-generation/split-canvas-editorial-art-poster/1/cover.jpg differdiff --git a/template/image-generation/street-scroll-backdrop-fashion-editorial/1/cover.jpg b/template/image-generation/street-scroll-backdrop-fashion-editorial/1/cover.jpgindex 7582c0ff..b38a5916 100644Binary files a/template/image-generation/street-scroll-backdrop-fashion-editorial/1/cover.jpg and b/template/image-generation/street-scroll-backdrop-fashion-editorial/1/cover.jpg differdiff --git a/template/image-generation/tech-product-review-thumbnail/1/cover.jpg b/template/image-generation/tech-product-review-thumbnail/1/cover.jpgindex 8acf28a8..1d42679b 100644Binary files a/template/image-generation/tech-product-review-thumbnail/1/cover.jpg and b/template/image-generation/tech-product-review-thumbnail/1/cover.jpg differdiff --git a/template/image-generation/travel-souvenir-enamel-pin/1/cover.jpg b/template/image-generation/travel-souvenir-enamel-pin/1/cover.jpgindex acf00f59..2f93e514 100644Binary files a/template/image-generation/travel-souvenir-enamel-pin/1/cover.jpg and b/template/image-generation/travel-souvenir-enamel-pin/1/cover.jpg differdiff --git a/template/image-generation/travel-souvenir-enamel-pin/1/template.json b/template/image-generation/travel-souvenir-enamel-pin/1/template.jsonindex e5f6c6bd..d9bdb4b9 100644--- a/template/image-generation/travel-souvenir-enamel-pin/1/template.json+++ b/template/image-generation/travel-souvenir-enamel-pin/1/template.json@@ -33,7 +33,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "Turn the reference photo into a travel souvenir enamel pin badge. Compose it as a SCENE, not a single isolated object.\n\nSubject hierarchy: the defining landscape, terrain or landmark of the photo forms the main body of the badge and occupies most of its area. If a person appears prominently in the photo, keep them in the badge as a small, simplified figure at true relative scale within that landscape — the person is an accent, the landscape is the subject. Preserve the original spatial relationship and scale between the figure and the surroundings.\n\nHow to render the person: flat enamel color blocks matching their real clothing and hair color from the photo. The face is a smooth plain area of light skin-tone enamel with no drawn facial features — do NOT render the person as a dark or black silhouette, and do NOT black out the face or head. Skin reads as a warm light enamel color, clearly lighter than the clothing.\n\nStyling: thin polished gold outline around the silhouette and along every internal divider, glossy enamel color fill, gentle even lighting with only a soft sheen on the gold lines, very subtle drop shadow. Outer contour follows the scene's own shape, not a plain rectangle.\n\nBackground: flat dark navy coarse linen texture. Badge centered, filling about 60% of the frame.\n\nAvoid: black silhouette figure, blacked-out face, dark featureless head, portrait close-up, detailed facial features, person dominating the badge, cropping out the landscape, three-quarter angle, macro product photography, heavy specular glare, cartoon, realistic scene, text, watermark."+ "prompt": "将参考照片变成旅行纪念珐琅徽章。将其组成一个场景,而不是单个孤立的对象。\n\n主题层次:照片的定义景观、地形或地标构成徽章的主体并占据其大部分区域。如果一个人出现在照片中的显着位置,请将其作为一个小而简化的人物以真实的相对比例保留在该风景中 - 人物只是点缀,风景才是主体。保留人物与周围环境原有的空间关系和尺度。\n\n如何渲染人物:平面珐琅色块与照片中的真实服装和头发颜色相匹配。脸部是浅肤色珐琅的平滑纯色区域,没有绘制面部特征 - 不要将人物渲染为深色或黑色轮廓,也不要遮蔽脸部或头部。皮肤呈现出温暖的浅珐琅色,明显比衣服浅。\n\n造型:轮廓周围和每个内部分隔线周围都有细细的抛光金色轮廓,有光泽的珐琅颜色填充,柔和均匀的灯光,金色线条上只有柔和的光泽,非常微妙的阴影。外部轮廓遵循场景自身的形状,而不是简单的矩形。\n\n背景:平坦的深海军蓝粗亚麻纹理。徽章居中,填充约 60% 的框架。\n\n避免:黑色剪影人物、涂黑的脸、深色无特征的头部、人像特写、详细的面部特征、人主宰徽章、裁剪风景、四分之三角度、微距产品摄影、强烈的镜面眩光、卡通、现实场景、文字、水印。"},"en-US": {"prompt": "Turn the reference photo into a travel souvenir enamel pin badge. Compose it as a SCENE, not a single isolated object.\n\nSubject hierarchy: the defining landscape, terrain or landmark of the photo forms the main body of the badge and occupies most of its area. If a person appears prominently in the photo, keep them in the badge as a small, simplified figure at true relative scale within that landscape — the person is an accent, the landscape is the subject. Preserve the original spatial relationship and scale between the figure and the surroundings.\n\nHow to render the person: flat enamel color blocks matching their real clothing and hair color from the photo. The face is a smooth plain area of light skin-tone enamel with no drawn facial features — do NOT render the person as a dark or black silhouette, and do NOT black out the face or head. Skin reads as a warm light enamel color, clearly lighter than the clothing.\n\nStyling: thin polished gold outline around the silhouette and along every internal divider, glossy enamel color fill, gentle even lighting with only a soft sheen on the gold lines, very subtle drop shadow. Outer contour follows the scene's own shape, not a plain rectangle.\n\nBackground: flat dark navy coarse linen texture. Badge centered, filling about 60% of the frame.\n\nAvoid: black silhouette figure, blacked-out face, dark featureless head, portrait close-up, detailed facial features, person dominating the badge, cropping out the landscape, three-quarter angle, macro product photography, heavy specular glare, cartoon, realistic scene, text, watermark."diff --git a/template/image-generation/vintage-scientific-botanical-poster/1/cover.jpg b/template/image-generation/vintage-scientific-botanical-poster/1/cover.jpgindex d822a7ef..5c6bf5fc 100644Binary files a/template/image-generation/vintage-scientific-botanical-poster/1/cover.jpg and b/template/image-generation/vintage-scientific-botanical-poster/1/cover.jpg differdiff --git a/template/image-generation/vr-headset-exploded-interface-diagram/1/cover.jpg b/template/image-generation/vr-headset-exploded-interface-diagram/1/cover.jpgindex b68f80af..dbb0b82f 100644Binary files a/template/image-generation/vr-headset-exploded-interface-diagram/1/cover.jpg and b/template/image-generation/vr-headset-exploded-interface-diagram/1/cover.jpg differdiff --git a/template/image-generation/watercolor-editorial-illustration-poster/1/cover.jpg b/template/image-generation/watercolor-editorial-illustration-poster/1/cover.jpgindex b957bdf8..9b7c9fec 100644Binary files a/template/image-generation/watercolor-editorial-illustration-poster/1/cover.jpg and b/template/image-generation/watercolor-editorial-illustration-poster/1/cover.jpg differdiff --git a/template/image-generation/weekly-outfit-infographic/1/cover.jpg b/template/image-generation/weekly-outfit-infographic/1/cover.jpgindex d43d0642..ed9db48f 100644Binary files a/template/image-generation/weekly-outfit-infographic/1/cover.jpg and b/template/image-generation/weekly-outfit-infographic/1/cover.jpg differdiff --git a/template/image-generation/wuhan-breakfast-food-map-infographic/1/cover.jpg b/template/image-generation/wuhan-breakfast-food-map-infographic/1/cover.jpgindex b6eab27c..caffda3e 100644Binary files a/template/image-generation/wuhan-breakfast-food-map-infographic/1/cover.jpg and b/template/image-generation/wuhan-breakfast-food-map-infographic/1/cover.jpg differdiff --git a/template/image-generation/wuhan-vintage-map-poster/1/cover.jpg b/template/image-generation/wuhan-vintage-map-poster/1/cover.jpgindex 7a44035b..9fa989c1 100644Binary files a/template/image-generation/wuhan-vintage-map-poster/1/cover.jpg and b/template/image-generation/wuhan-vintage-map-poster/1/cover.jpg differdiff --git a/template/image-generation/y2k-streetwear-mobile-landing-page/1/cover.jpg b/template/image-generation/y2k-streetwear-mobile-landing-page/1/cover.jpgindex 1eca1973..45535244 100644Binary files a/template/image-generation/y2k-streetwear-mobile-landing-page/1/cover.jpg and b/template/image-generation/y2k-streetwear-mobile-landing-page/1/cover.jpg differdiff --git a/template/image-generation/y2k-virtual-try-on-game-ui/1/cover.jpg b/template/image-generation/y2k-virtual-try-on-game-ui/1/cover.jpgindex b7757f21..ec96020e 100644Binary files a/template/image-generation/y2k-virtual-try-on-game-ui/1/cover.jpg and b/template/image-generation/y2k-virtual-try-on-game-ui/1/cover.jpg differdiff --git a/template/video-generation/beach-stone-stop-motion/1/cover.jpg b/template/video-generation/beach-stone-stop-motion/1/cover.jpgindex 2c39ac5f..37fd84f0 100644Binary files a/template/video-generation/beach-stone-stop-motion/1/cover.jpg and b/template/video-generation/beach-stone-stop-motion/1/cover.jpg differdiff --git a/template/video-generation/bedroom-falling-book-catch/1/cover.jpg b/template/video-generation/bedroom-falling-book-catch/1/cover.jpgindex 03123131..03c0267a 100644Binary files a/template/video-generation/bedroom-falling-book-catch/1/cover.jpg and b/template/video-generation/bedroom-falling-book-catch/1/cover.jpg differdiff --git a/template/video-generation/first-person-dragon-rider/1/cover.jpg b/template/video-generation/first-person-dragon-rider/1/cover.jpgindex 1dc078c6..d55e389b 100644Binary files a/template/video-generation/first-person-dragon-rider/1/cover.jpg and b/template/video-generation/first-person-dragon-rider/1/cover.jpg differdiff --git a/template/video-generation/japanese-influencer-get-ready-ugc/1/cover.jpg b/template/video-generation/japanese-influencer-get-ready-ugc/1/cover.jpgindex 30aebebd..8f28245c 100644Binary files a/template/video-generation/japanese-influencer-get-ready-ugc/1/cover.jpg and b/template/video-generation/japanese-influencer-get-ready-ugc/1/cover.jpg differdiff --git a/template/video-generation/luxury-jewelry-unboxing/1/cover.jpg b/template/video-generation/luxury-jewelry-unboxing/1/cover.jpgindex d54e8817..a966f9df 100644Binary files a/template/video-generation/luxury-jewelry-unboxing/1/cover.jpg and b/template/video-generation/luxury-jewelry-unboxing/1/cover.jpg differdiff --git a/template/video-generation/rainforest-mysterious-light/1/cover.jpg b/template/video-generation/rainforest-mysterious-light/1/cover.jpgindex 78ed4834..94f6cb73 100644Binary files a/template/video-generation/rainforest-mysterious-light/1/cover.jpg and b/template/video-generation/rainforest-mysterious-light/1/cover.jpg differdiff --git a/template/video-generation/recipe-infographic-cooking-sequence/1/cover.jpg b/template/video-generation/recipe-infographic-cooking-sequence/1/cover.jpgindex d339e6b5..898bb21b 100644Binary files a/template/video-generation/recipe-infographic-cooking-sequence/1/cover.jpg and b/template/video-generation/recipe-infographic-cooking-sequence/1/cover.jpg differdiff --git a/template/video-generation/recipe-infographic-cooking-sequence/1/template.json b/template/video-generation/recipe-infographic-cooking-sequence/1/template.jsonindex 1b71d9ee..9373bf19 100644--- a/template/video-generation/recipe-infographic-cooking-sequence/1/template.json+++ b/template/video-generation/recipe-infographic-cooking-sequence/1/template.json@@ -32,7 +32,7 @@},"defaultsByLocale": {"zh-CN": {- "prompt": "MODEL: Gemini Omni 1.1 Flash\n\nSHOT STRUCTURE: 5 shots, 10 seconds, vertical 9:16, exactly as listed.\n\nREFS:\n\n@Image1 = visual storyboard reference. It controls the cooking order, black cast-iron skillet, ingredients, finished-dish styling, dark surfaces, and soft window-lit food photography. Convert its picture panels into fresh full-frame live-action cooking shots.\n\nVISUAL CONTENT LOCK:\n\nEvery frame contains only food, cookware, ingredients, natural hands, and kitchen surfaces. Live-action cooking imagery fills the entire 9:16 frame. All surfaces remain clean and unlettered. The source poster’s typography, grid, headings, step numbers, captions, logos, and graphic elements remain reference metadata outside the generated video.\n\nAUDIO CONTENT LOCK:\n\nThe soundtrack consists exclusively of synchronized cooking sounds and quiet kitchen room tone. Close dry chopping, oil pouring, pan sizzling, wooden-spoon movement, vegetables landing, tomato pouring, bubbling sauce, and soft steam. Spoken voices, narration, vocals, and music remain outside the soundtrack.\n\nGLOBAL LOOK:\n\n70mm large-format film aesthetic, fine organic grain, broad dynamic range, deep food detail, soft highlight roll-off, restrained anamorphic character. Vertical 9:16 composition with the food, cookware, and hands inside the central safe area. Soft diffused daylight enters from camera-left through a white curtain; pale stone counters return gentle fill. Saturated tomato red contrasts with black cast iron and warm neutral kitchen surfaces. Preserve realistic steam, oil flow, translucent onion, red-pepper texture, spice granules, sauce bubbles, and natural hand movement.\n\nSHOT 1 (0–1.5s) — Finished-dish hook\n\ncamera: three-quarter close hero view, quick smooth push toward the skillet.\n\naction_visual: finished shakshuka bubbles gently beside crusty bread; steam crosses the soft window light.\n\nsound: close bubbling sauce and faint skillet sizzle.\n\nexit: the circular skillet rim fills the frame.\n\n(MATCH CUT TO)\n\nSHOT 2 (1.5–3.5s) — Prepare ingredients\n\ncamera: locked overhead medium shot.\n\naction_visual: hands finish dicing onion and slicing red pepper, then arrange minced garlic, spices, tomatoes, eggs, feta, and herbs in small bowls.\n\nsound: clean knife rhythm, cutting-board taps, bowls touching the stone counter.\n\nexit: one hand lifts an olive-oil bottle.\n\n(CUTTING ON ACTION TO)\n\nSHOT 3 (3.5–5.5s) — Heat oil and sauté onion\n\ncamera: macro three-quarter insert shifting into an overhead close shot.\n\naction_visual: golden olive oil pours into the black skillet; diced onion follows immediately and a wooden spoon stirs through the rising sizzle.\n\nsound: smooth oil pour followed by a sharp fresh sizzle and dry wooden scraping.\n\nexit: the spoon sweeps screen-right.\n\n(MATCH ON MOVEMENT TO)\n\nSHOT 4 (5.5–7.5s) — Add pepper, garlic, and spices\n\ncamera: overhead close shot with a subtle push-in.\n\naction_visual: red pepper strips fall into the softened onion, followed by minced garlic, cumin, smoked paprika, and chili flakes. The spoon folds everything together in one continuous action.\n\nsound: vegetables landing, brighter pan sizzle, granular spices brushing the skillet.\n\nexit: paprika-red oil spreads beneath the spoon.\n\n(CUT TO)\n\nSHOT 5 (7.5–10s) — Add tomatoes and simmer\n\ncamera: low three-quarter close view, short dolly back.\n\naction_visual: crushed tomatoes pour into the skillet. A clean cooking-time jump reveals the sauce bubbling, reducing, and becoming visibly thicker.\n\nsound: dense tomato pour transitioning into wet clustered bubbling.\n\nexit: the thick bubbling red surface fills the final frame; bubbling continues across the extension."+ "prompt": "模型:Gemini Omni 1.1 Flash\n\n镜头结构:严格按照下列顺序生成 5 个镜头,总时长 10 秒,竖屏 9:16。\n\n参考图:\n\n@Image1 是视觉故事板参考,用于约束烹饪顺序、黑色铸铁煎锅、食材、成品造型、深色操作台面和柔和窗光下的食物摄影风格。将参考图中的分格画面转化为全新的、铺满画幅的实拍烹饪镜头。\n\n画面内容锁定:\n\n每一帧只出现食物、厨具、食材、自然入镜的双手和厨房台面。实拍烹饪画面必须铺满整个 9:16 画幅。所有画面区域保持干净,不出现文字。源海报中的排版、网格、标题、步骤编号、说明文字、标志及其他图形元素仅作为参考信息,不得出现在生成的视频中。\n\n声音内容锁定:\n\n音轨只包含与画面同步的烹饪声音和安静的厨房环境底噪,包括近距离切菜声、倒油声、煎锅滋滋声、木勺移动声、蔬菜落锅声、倒入番茄声、酱汁冒泡声和轻柔蒸汽声。不得出现对白、旁白、人声或音乐。\n\n整体视觉:\n\n70mm 大画幅电影质感,细腻自然的胶片颗粒,宽广动态范围,丰富食物细节,柔和的高光过渡和克制的变形宽银幕特征。采用竖屏 9:16 构图,食物、厨具和双手保持在中央安全区。柔和漫射日光从画面左侧穿过白色窗帘进入,浅色石材台面提供温柔补光。饱和番茄红与黑色铸铁和温暖中性的厨房表面形成对比。保留真实的蒸汽、油流、半透明洋葱、红椒纹理、香料颗粒、酱汁气泡和自然手部动作。\n\n镜头 1(0–1.5 秒):成品钩子镜头\n相机:四分之三近景英雄视角,快速、平滑地推向煎锅。\n画面动作:完成的北非蛋(shakshuka)在硬皮面包旁轻轻冒泡,蒸汽穿过柔和窗光。\n声音:近距离酱汁冒泡声和轻微煎锅滋滋声。\n收尾画面:圆形煎锅边缘充满画面。\n转场:匹配剪辑至下一镜头。\n\n镜头 2(1.5–3.5 秒):准备食材\n相机:锁定机位的俯拍中景。\n画面动作:双手完成洋葱切丁和红椒切片,随后把蒜末、香料、番茄、鸡蛋、菲达奶酪和香草分别摆入小碗。\n声音:干净利落的切刀节奏、刀接触砧板的敲击声,以及小碗接触石材台面的声音。\n收尾画面:一只手拿起橄榄油瓶。\n转场:通过动作衔接剪辑进入下一镜头。\n\n镜头 3(3.5–5.5 秒):热油并炒洋葱\n相机:微距四分之三插入镜头,随后转为俯拍特写。\n画面动作:金色橄榄油倒入黑色煎锅,紧接着加入洋葱丁,木勺在不断增强的滋滋声中搅拌。\n声音:顺滑倒油声,随后是清脆的新鲜滋滋声和木勺刮过锅底的干涩声。\n收尾画面:木勺向画面右侧扫过。\n转场:匹配运动转场至下一镜头。\n\n镜头 4(5.5–7.5 秒):加入红椒、蒜和香料\n相机:俯拍近景,轻微向前推进。\n画面动作:红椒条落入变软的洋葱,随后加入蒜末、小茴香、烟熏红椒粉和辣椒碎。木勺以一个连续动作把所有食材翻拌均匀。\n声音:蔬菜落锅声、更明亮的煎锅滋滋声,以及颗粒状香料擦过锅面的声音。\n收尾画面:被红椒粉染红的油从木勺下方扩散。\n转场:直接切至下一镜头。\n\n镜头 5(7.5–10 秒):加入番茄并炖煮\n相机:低机位四分之三近景,短距离滑轨后移。\n画面动作:压碎的番茄倒入煎锅。通过清晰的烹饪时间跳切,表现酱汁持续冒泡、收汁并明显变稠。\n声音:浓稠番茄倒入声逐渐过渡为湿润、密集的冒泡声。\n收尾画面:浓稠、冒泡的红色酱汁铺满最后一帧,并持续冒泡至镜头结束。"},"en-US": {"prompt": "MODEL: Gemini Omni 1.1 Flash\n\nSHOT STRUCTURE: 5 shots, 10 seconds, vertical 9:16, exactly as listed.\n\nREFS:\n\n@Image1 = visual storyboard reference. It controls the cooking order, black cast-iron skillet, ingredients, finished-dish styling, dark surfaces, and soft window-lit food photography. Convert its picture panels into fresh full-frame live-action cooking shots.\n\nVISUAL CONTENT LOCK:\n\nEvery frame contains only food, cookware, ingredients, natural hands, and kitchen surfaces. Live-action cooking imagery fills the entire 9:16 frame. All surfaces remain clean and unlettered. The source poster’s typography, grid, headings, step numbers, captions, logos, and graphic elements remain reference metadata outside the generated video.\n\nAUDIO CONTENT LOCK:\n\nThe soundtrack consists exclusively of synchronized cooking sounds and quiet kitchen room tone. Close dry chopping, oil pouring, pan sizzling, wooden-spoon movement, vegetables landing, tomato pouring, bubbling sauce, and soft steam. Spoken voices, narration, vocals, and music remain outside the soundtrack.\n\nGLOBAL LOOK:\n\n70mm large-format film aesthetic, fine organic grain, broad dynamic range, deep food detail, soft highlight roll-off, restrained anamorphic character. Vertical 9:16 composition with the food, cookware, and hands inside the central safe area. Soft diffused daylight enters from camera-left through a white curtain; pale stone counters return gentle fill. Saturated tomato red contrasts with black cast iron and warm neutral kitchen surfaces. Preserve realistic steam, oil flow, translucent onion, red-pepper texture, spice granules, sauce bubbles, and natural hand movement.\n\nSHOT 1 (0–1.5s) — Finished-dish hook\n\ncamera: three-quarter close hero view, quick smooth push toward the skillet.\n\naction_visual: finished shakshuka bubbles gently beside crusty bread; steam crosses the soft window light.\n\nsound: close bubbling sauce and faint skillet sizzle.\n\nexit: the circular skillet rim fills the frame.\n\n(MATCH CUT TO)\n\nSHOT 2 (1.5–3.5s) — Prepare ingredients\n\ncamera: locked overhead medium shot.\n\naction_visual: hands finish dicing onion and slicing red pepper, then arrange minced garlic, spices, tomatoes, eggs, feta, and herbs in small bowls.\n\nsound: clean knife rhythm, cutting-board taps, bowls touching the stone counter.\n\nexit: one hand lifts an olive-oil bottle.\n\n(CUTTING ON ACTION TO)\n\nSHOT 3 (3.5–5.5s) — Heat oil and sauté onion\n\ncamera: macro three-quarter insert shifting into an overhead close shot.\n\naction_visual: golden olive oil pours into the black skillet; diced onion follows immediately and a wooden spoon stirs through the rising sizzle.\n\nsound: smooth oil pour followed by a sharp fresh sizzle and dry wooden scraping.\n\nexit: the spoon sweeps screen-right.\n\n(MATCH ON MOVEMENT TO)\n\nSHOT 4 (5.5–7.5s) — Add pepper, garlic, and spices\n\ncamera: overhead close shot with a subtle push-in.\n\naction_visual: red pepper strips fall into the softened onion, followed by minced garlic, cumin, smoked paprika, and chili flakes. The spoon folds everything together in one continuous action.\n\nsound: vegetables landing, brighter pan sizzle, granular spices brushing the skillet.\n\nexit: paprika-red oil spreads beneath the spoon.\n\n(CUT TO)\n\nSHOT 5 (7.5–10s) — Add tomatoes and simmer\n\ncamera: low three-quarter close view, short dolly back.\n\naction_visual: crushed tomatoes pour into the skillet. A clean cooking-time jump reveals the sauce bubbling, reducing, and becoming visibly thicker.\n\nsound: dense tomato pour transitioning into wet clustered bubbling.\n\nexit: the thick bubbling red surface fills the final frame; bubbling continues across the extension."