Prompt library
PROMPTS
1328 hand-tested AI prompts across image, video, and music — each with full context, variables, and a sample output. Content is authored in Traditional Chinese; the UI here is English.
02Tool (Image)
03Use case
Page 10/10(221 total)
IMAGE
免費01IMAGE／GPT Image 2
2026-04-22
GPT Image 2 infographics: tell the whole data story in one image
GPT Image 2's most underrated ability is making infographics. Its Chinese character rendering is 10x more reliable than Midjourney's; feed it data plus structure and it produces an infographic ready for an annual report, a deck, or LinkedIn.
SAMPLE一張 1024×1024 資訊圖表：米白背景、主標題「2026 台灣 AI 產業就業分布」用粗黑體放左上；中央是一個簡潔的水平長條圖，四個數據 call-out 分佈在四個角落 — 「47%：軟體工程師」「22%：資料科學家」「18%：產品經理」「13%：研究與學術」；底部一行小
IntermediateGen in-site
Read
IMAGE
PRO02IMAGE／GPT Image 2
2026-04-22
GPT Image 2 Chinese-typography posters: no garbled fonts
The biggest pain point with Chinese posters is that every image model garbles the characters. GPT Image 2 currently renders Traditional Chinese better than anything else, and this prompt is built for "Chinese headline + layout + zero errors," print-ready out of the box.
SAMPLE一張 1024×1536 中文活動海報：米白背景、上三分之一放粗黑體主標題「新銳設計 春季選展」兩行分列、字距 tight；副標題「2026 / 04 / 28 — 松菸 · 台北文創」細黑體置於主標下方；下半部是一個朱紅色圓角三角幾何圖形抽象元素，右下方資訊欄放日期地點贊助 l
IntermediateGen in-site
Read
PRO03IMAGE／Flux.1 Pro
2026-04-22
Flux.1 Fashion Lookbook: Launch a Small Brand's Visuals in One Week
A single prompt produces a full visual set: model shots, product shots, street style, and action shots — with consistent color, lighting, and text placement throughout.
SAMPLE一整組 6 張、色溫一致的編輯式拍攝：F1 主視覺是混血亞裔模特兒穿著米白亞麻外套站在水泥牆前；F2 是疊好的針織與襯衫平鋪在石板上；F3 麻布纖維特寫看得到織紋；F4 模特兒在里斯本巷弄裡邊走邊回頭；F5 站在極簡建築前的全身編輯照；F6 雙手捧著皮革手環的特寫。沙色 + 骨白
IntermediateGen in-site
Read
