{ "image_prompt": { "subject": { "description": "年轻女性,肤色白皙,长直发,发色为 {argument name=\"hair color\" default=\"浅棕色\"},部分发丝随风飘动。", "pose": "自拍角度,头部略微倾斜,{argument name=\"expression\" default=\"俏皮地吐着舌头\"}," } } }
[Subject]: ____ [Chinese Main Title]: ____ [English Title]: ____ [Central Subject]: ____ [Corner Pressure Method]: Four-corner text pressure / Dual-corner diagonal pressure / Triple-corner wrap + one corner release / Single-corner heavy pre

Half-sleeved palace dress; clear shoulder lines; fabric naturally supported by the body shape; large area of skin exposed on the chest, slender waist is evident; clear waist-to-hip ratio; picking hairpins at a flower market; bowing head to

Use the attached reference image. Preserve her facial identity 100% — same face shape, eyes, eyebrows, nose, lips, skin tone, hair color, hairstyle, and natural expression. Remove the glasses while keeping her appearance realistic and insta

{ "base_image": "上传的图片", "style_transfer": { "visual_characteristics": { "head_and_face": { "material": "半透明树脂,内嵌星光和发光神经电路", "surface_effect": "镜面高光效果,带有金色丝线纹理和银河般的反射",

一张高分辨率的照片,展示了一个透明塑料抽绳袋放置在浅灰色背景上。袋子内部整齐地摆放着多个微型3D [主题]人偶。袋子用柔软的白色丝带系紧,并附有一个黑色标签,上面写着“[标签文字]”。柔和的灯光和清晰的阴影突出了逼真的纹理和细节。

9:16 vertical version, hyper-realistic real-life photography × Dunhuang Feitian mythology × Eastern female warrior × High-end cinematic Editorial × Dynamic maximalism aesthetics. Using strong asymmetrical composition and large-depth dynamic

Generate a 9:16 "Torn Paper Narrative Editorial Poster". Main Scene: [Fill in character / scene] Main Visual Medium: [Modern fashion photography / Contemporary documentary photography, etc.] Hidden Image: [Fill in another layer of story un

Theme direction: Oriental Zen minimalist cover poster Style branch: High-viral conceptual type Subject content: An ancient-style woman standing on the edge of a light-colored platform, looking sideways at a vertically falling strip of light

Create a premium 4:5 vertical architectural travel artwork for [COUNTRY / CITY / LOCATION / LANDMARK]. TEXT-TO-IMAGE ONLY — generate the entire scene creatively from [COUNTRY / CITY / LOCATION / LANDMARK] alone. No reference image required