Testing Strict Prompt Adherence in GPT Image 1.5 vs. Gemini

このAI画像プロンプトについて

The user compares GPT Image 1.5 and Gemini using the same detailed prompt, finding that GPT Image 1.5 is more 'creative' and less adherent to strict rules, while Gemini shows better compliance. The prompt details specific visual constraints for an isometric city wall scene. - YouMind

A vertical split-screen creative product photography composition on a clean white wall background. High-resolution, photorealistic, commercial advertisement quality.

Top Section: The Specimen Box
The upper half features an exquisite light oak wooden shadow box frame mounted on the wall. Inside, a specific outfit is displayed as an artistic flat-lay museum specimen: {argument name="Clothing Details" default="a sleek black satin slip dress with delicate lace trim and thin spaghetti straps"}. The garments are neatly pinned in place. Surrounding them are small thematic decorative props: {argument name="Props" default="dried roses, vintage perfume bottles, silk ribbon"}. Elegant calligraphy on the matte paper backdrop reads: {argument name="Text" default="Midnight Elegance"}. Soft studio lighting accentuates the rich texture and drape of the fabric.

Bottom Section: Naked-Eye 3D Reality
The lower half creates a hyperrealistic "naked-eye 3D" illusion. A rectangular picture-frame border sits directly beneath the top box. A stunningly realistic young woman {argument name="Model Description" default="a poised East Asian model with long wavy black hair, subtle smoky eyes, and a confident gaze"} wears the exact same outfit as shown above.

She lounges casually on the bottom edge of the frame—one leg bent with foot resting inside the frame, the other leg elegantly dangling out into the viewer’s space. Her torso leans back slightly, elbow resting on her raised knee, fingers lightly grazing the fabric near her collarbone. Her body forms a soft, sensual S-curve that highlights the garment’s silhouette without overt exposure. She looks directly at the camera with a calm, knowing smile—inviting yet enigmatic. This dynamic, lifelike pose contrasts powerfully with the static, archival display above, creating visual tension between reality and presentation.

Technical Specs:
Soft natural shadows, ambient occlusion, bright and airy yet cinematic lighting, 8K resolution, Octane Render, vivid but refined color palette, ultra-detailed fabric textures (satin sheen, lace transparency, stitching), shallow depth of field, Vogue editorial style, filmic grain, professional fashion photography.

Negative Prompt (recommended):
blurry, low-res, distorted anatomy, extra limbs, deformed hands, cartoon, anime, doll-like, plastic skin, overexposed, cluttered background, text errors, mismatched clothing, floating objects, unrealistic proportions.73:T99d,

このAI画像プロンプトテンプレートの使い方

1. 1プロンプトをコピー — テンプレートのプロンプトとネガティブプロンプトを取得。
2. 2モデルを選ぶ — 最適な推奨AIモデルを選択。
3. 3生成 — ワンクリックでスタジオを開いて作成。