{"id":"32660","prompt":"Goal: Create a photorealistic 2x2 summer festival vlog collage featuring one cute mascot robot character, a lily-of-the-valley fairy robot, enjoying Japanese matsuri street food at night.\n\nCanvas: Vertical 9:16 image, divided into exactly 4 equal panels by clean white borders, with warm cinematic night photography, shallow depth of field, realistic lens bokeh, and glowing paper lanterns.\n\nCharacter details: The same single character appears in every panel: a glossy white-and-green chibi robot with a round face, huge sparkling blue eyes with star highlights, pink blush cheeks, small smiling mouth, green leaf-shaped ears, green joints and boots, and a curved green stem sprouting from the head with exactly 2 small white bell-shaped lily-of-the-valley flowers. The character wears a pale cream yukata with delicate green floral patterns and a green obi sash. Keep the character toy-like and polished, with realistic reflections and soft festival lighting.\n\nLayout and panel content: Use exactly 4 discrete scenes. Panel 1, top left: full-body walking shot down a festival street, the character happily holding and biting a chocolate-coated banana with sprinkles; red food stall awnings and lanterns behind. Panel 2, top right: seated close-up on a wooden bench, eating yakisoba noodles in a wooden tray with chopsticks, noodles lifted toward the mouth. Panel 3, bottom left: standing close-up, eyes squeezed shut in delight while eating pink-and-white shaved ice in a paper cup with a small spoon. Panel 4, bottom right: rear three-quarter view of the character walking away through the lantern-lit crowd, turning slightly with a cute satisfied smile; no food in hand.\n\nBackground: A lively Japanese summer festival at night with blurred people in yukata, cobblestone path, warm orange lantern strings, food stalls, red banners, soft haze, and rich golden highlights. The setting should feel candid, joyful, and immersive like a travel vlog thumbnail.\n\nVisual style: High-end photorealistic 3D mascot photography, ultra-detailed materials, glossy ceramic/plastic surfaces, realistic fabric texture on the yukata, warm color grading, cinematic depth of field, vertical social-media collage composition.\n\nConstraints: Include exactly 1 recurring character, exactly 4 panels, exactly 3 visible food items across the first three panels, and exactly 2 white bell flowers on the head stem in every panel. Do not add text, captions, logos, watermarks, speech bubbles, or extra characters in focus.","promptEn":"Goal: Create a photorealistic 2x2 summer festival vlog collage featuring one cute mascot robot character, a lily-of-the-valley fairy robot, enjoying Japanese matsuri street food at night.\n\nCanvas: Vertical 9:16 image, divided into exactly 4 equal panels by clean white borders, with warm cinematic night photography, shallow depth of field, realistic lens bokeh, and glowing paper lanterns.\n\nCharacter details: The same single character appears in every panel: a glossy white-and-green chibi robot with a round face, huge sparkling blue eyes with star highlights, pink blush cheeks, small smiling mouth, green leaf-shaped ears, green joints and boots, and a curved green stem sprouting from the head with exactly 2 small white bell-shaped lily-of-the-valley flowers. The character wears a pale cream yukata with delicate green floral patterns and a green obi sash. Keep the character toy-like and polished, with realistic reflections and soft festival lighting.\n\nLayout and panel content: Use exactly 4 discrete scenes. Panel 1, top left: full-body walking shot down a festival street, the character happily holding and biting a chocolate-coated banana with sprinkles; red food stall awnings and lanterns behind. Panel 2, top right: seated close-up on a wooden bench, eating yakisoba noodles in a wooden tray with chopsticks, noodles lifted toward the mouth. Panel 3, bottom left: standing close-up, eyes squeezed shut in delight while eating pink-and-white shaved ice in a paper cup with a small spoon. Panel 4, bottom right: rear three-quarter view of the character walking away through the lantern-lit crowd, turning slightly with a cute satisfied smile; no food in hand.\n\nBackground: A lively Japanese summer festival at night with blurred people in yukata, cobblestone path, warm orange lantern strings, food stalls, red banners, soft haze, and rich golden highlights. The setting should feel candid, joyful, and immersive like a travel vlog thumbnail.\n\nVisual style: High-end photorealistic 3D mascot photography, ultra-detailed materials, glossy ceramic/plastic surfaces, realistic fabric texture on the yukata, warm color grading, cinematic depth of field, vertical social-media collage composition.\n\nConstraints: Include exactly 1 recurring character, exactly 4 panels, exactly 3 visible food items across the first three panels, and exactly 2 white bell flowers on the head stem in every panel. Do not add text, captions, logos, watermarks, speech bubbles, or extra characters in focus.","promptZh":"目标：创作一张照片级写实风格的 2x2 夏季庆典 Vlog 拼图，主角为一个可爱的吉祥物机器人，铃兰仙子机器人，正在夜色中享用日本祭典街头美食。\n\n画布：垂直 9:16 图像，由整洁的白色边框精确划分为 4 个等大的面板，采用温暖的电影感夜景摄影，具有浅景深、逼真的镜头光斑和发光的纸灯笼。\n\n角色细节：每个面板中均出现同一个角色：一个光泽感十足的白绿配色 Q 版机器人，拥有圆脸、闪烁着星光的大蓝眼睛、粉色腮红、微笑着的小嘴、叶片状的绿色耳朵、绿色关节和靴子，头顶长出一根弯曲的绿色茎，上面精确地开着 2 朵白色铃铛状铃兰花。角色身穿淡奶油色浴衣，上面有精致的绿色花卉图案和绿色腰带。保持角色玩具般的质感和精致度，呈现逼真的反射效果和柔和的庆典灯光。\n\n布局与面板内容：使用 4 个独立的场景。面板 1（左上）：在庆典街道上行走的全身镜头，角色开心地拿着并咬着 巧克力香蕉配糖针；背景是红色的摊位遮阳棚和灯笼。面板 2（右上）：坐在木凳上的特写，正用筷子吃着 木盒装日式炒面，面条正被夹起送往嘴边。面板 3（左下）：站立特写，正用小勺吃着 纸杯装粉白刨冰，眼睛因美味而眯起。面板 4（右下）：角色背影的三分之四侧面视角，正走在灯火通明的人群中，微微回头露出满足的可爱微笑；手中没有食物。\n\n背景：热闹的日本夏季夜间庆典，背景有穿着浴衣的人群虚影、鹅卵石小径、温暖的橙色灯笼串、美食摊位、红色旗帜、柔和的雾气和丰富的金色高光。场景应具有抓拍感、欢乐感，并像旅行 Vlog 缩略图一样具有沉浸感。\n\n视觉风格：高端照片级写实 3D 吉祥物摄影，超精细材质，光泽感陶瓷/塑料表面，浴衣具有逼真的织物纹理，暖色调调色，电影级景深，垂直社交媒体拼图构图。\n\n约束条件：包含 1 个固定角色，精确 4 个面板，前三个面板中总共出现 3 种可见食物，且每个面板中头顶的茎上必须精确包含 2 朵白色铃铛花。不要添加任何文字、标题、Logo、水印、气泡或额外的焦点角色。"}