分格與分鏡腳本

四格寵物漫畫單圖

用一張靜態圖片排出同一隻寵物的四格漫畫,毛色以照片或文字為準並保持不變

更新日期:

完整提示詞

英文提示詞包含建議的範例需求。生成前請調整主體和細節。原始範例輸入及生成參數尚未完全核實。

推薦模型: GPT Image 2.5 Flare

Create one still image laid out as exactly four comic panels stacked in one column, with equal panel heights, even gutters, and a narrow outer margin. This deliverable is a static sheet, with no GIF, video, or animation export. Read the user request for the animal, its markings, and the action of each panel from top to bottom. If a pet photo is uploaded, keep species, coat pattern, ear shape, and collar identical in every panel, and do not turn the face human. If no photo is supplied, design one original animal from the description and repeat that design without drift. Give each panel one clear story beat and a readable setting, using the same line weight and palette throughout. Keep the animal large enough to identify. If the user supplies a short caption or spoken line for a panel, render that text a single time in the given spelling, inside a simple box or balloon. If no words are supplied, leave the sheet free of lettering. Do not add a fifth panel, a logo, or a watermark. Check that the sequence reads top to bottom and that markings, collar, and body proportions stay fixed while the background may change. User request: A cream rabbit with one gray ear moves through four beats: it studies a closed cardboard box on a kitchen floor, lifts the flap with one paw, sits inside among paper scraps, and finally leaves only the gray ear above the rim. No captions.

提示詞技巧

  • 數清楚正好四格,並確認毛色、耳朵和項圈沒有走樣
  • 如果寫了對白,逐格核對文字,確認沒有多餘字母或第五格

來源與署名

由 OmniArt 改編及重寫,以下列出來源署名、修改說明及使用條件

  • Four-panel pet comic · OpenAI

    RaphaelAI-OpenLab/awesome-gpt-image-2.6

    Public example from OpenAI's image prompting guide; no redistribution license was verified and the collection's MIT license does not cover it. This template is independently authored; no source text or media is copied.

    Replaced the fixed gag with a user-described four-panel still sheet, added pet-identity controls, and stated that no GIF or animation export is promised. The source collection labels this case GPT Image 2.6; GPT Image 2.5 Flare is a recommendation, not a verified example model.

來源與署名

相關的提示詞配方