Which AI model works best for this prompt?

This prompt is optimized for Sora 2, but it also works well in Midjourney and DALL·E for similar magical realism results.

Can I modify the text or colors?

Yes, you can easily customize the city name, date, color palette, or any other element to suit your project.

What resolution will the output be?

Resolution depends on the AI generator, but it's typically high-quality enough for digital posters, social media, and prints.

GPT Image 2AI 提示词

Guangzhou Goddess City Poster — Sora 2 Prompt

Create a magical vertical poster blending a graceful goddess with Guangzhou's illuminated skyline, perfect for artistic and promotional use.

提示词 · EN

Create a vertical fantasy city poster for Guangzhou, China. At the bottom of the composition, a beautiful ethereal woman with flowing white-golden hair holds a bouquet of colorful flowers, eyes closed peacefully. Her long hair transforms into swirling golden energy ribbons that rise upward, carrying glowing particles and sparkles. The golden streams merge into the illuminated Guangzhou skyline at the top, featuring the Canton Tower and modern skyscrapers glowing warmly against a deep black-red night sky. Add the text "广州·中国" and "2026/04/20" in elegant white typography at the upper right, with "LIYUE" as a small credit below. Rich warm golden palette with deep crimson and black contrast, magical realism style, cinematic lighting, vertical 9:16 ratio.

关于这个提示词

Create a magical vertical poster blending a graceful goddess with Guangzhou's illuminated skyline, perfect for artistic and promotional use. 它更适合作为 GPT Image 2 的概念艺术起点：先保留画面结构、主体关系和风格约束，再替换成你的品牌、人物或场景。

常见用途包括 posters、advanced、city poster、fantasy illustration。这些标签只作为创作方向参考，实际生成时仍应以提示词结构和预览图为准。

使用时建议先小幅修改主体、场景、镜头和色调，再生成多个版本对比构图与细节。这样页面内容对用户有实际帮助，也避免把模型名或标签机械堆在正文里。