A scene where 【Tokyo Tower】is occupied by a super gigantic, adorable 【cat】.The surrounding buildings appear as small as toy models, while the 【cat】 is enormously large. The setting features a realistic city environment. The overall mood is quiet, warm, soothing, and cute. 画面中,【东京塔】被一只超级巨大、超级可爱的【猫】占据。周围的建筑物看起来就像玩具模型一样小,而【猫】则非常巨大。 游戏背景设定在一个逼真的城市环境中。 整体氛围安静、温暖、舒缓、可爱。

Create a character design sheet based on the uploaded image. Design: Character profile poster Including: 1. Large main portrait 2. Various facial expressions (happy, angry, sleepy, surprised, etc.) 3. Action poses (running, playing) (base

Using the uploaded image as the exact visual base, transform it into a hyper-realistic 3D object that retains the original shape and proportions of the logo only. Apply traditional Ottoman Iznik ceramic textures—featuring a warm white glaze

Create a full-body 9:16 realistic fashion editorial featuring an adult young East Asian woman with a refined K-pop-inspired look, expressive eyes, a slim silhouette, and long balanced legs. Use the uploaded image as a pose reference only.

Create an extremely detailed portrait poster in a bold Japanese graffiti-inspired art style, blending modern urban street aesthetics with expressive Japanese visual culture. The poster should include dynamic graffiti typography, layered spr

{ "image_generation_request": { "content": { "subject_details": { "identity": { "name": "Ana de Armas", "gender": "female" }, "appearance": { "hair": { "color": "dark brown", "texture": "straight", "length": "long" }, "expression": "smiling

{ "project_type": "image_generation_from_reference", "preservation_constraints": { "facial_fidelity": "100% identical to the original image (features, bone structure, skin tone)", "expression_lock": true, "texture_quality": "Preserve realis

"Restore this old photo into a professional DSLR-grade portrait - high-quality color and detail, using advanced magnification algorithms, with results comparable to a Canon EOS R6 II. Ensure the restored image looks natural, retains accurat

Generate the text "◯%" inside a heart symbol above the head of each subject. The number inside ◯ should be a random number between 0 and 100.

Please generate an image where the person from the first image winks in the second image.