GPT Image 2.5 Prompts

One Line to X Carousel

X multi-image posts now scroll horizontally: the first image appears in full, while only part of the second is visible, prompting readers to swipe left to see what follows. This skill helps you create a “pull-apart” continuous four-image sequence—four vertical images combine into one complete landscape image, and each swipe reveals another section to the right. The right edge of the first image becomes the hook that makes readers want to swipe. Three input types are supported: · One-line idea → text-to-image generation, creating a continuous four-section wide image from scratch · One image → crop it directly if possible; otherwise, use it as a reference to reconstruct four continuous scenes · 2–4 images → automatically arrange the narrative order and blend them into one stylistically consistent continuous wide image Workflow: First, you’ll review a four-panel narrative script (hook → reveal → twist → ending). Then the skill generates a 21:9 wide image and uses the companion free public image slicer to precisely cut it into 4 PNGs locally in your browser (your images are never uploaded to any server). Finally, it provides the posting order and tweet copy. The default output ratio is 4:5, with 9:16 and 1:1 also supported. No text is placed inside the images. Ideal for creators who want their X multi-image posts to inspire more “page-turning” curiosity, as well as teams creating product, story, educational, or travel content. Only want to slice an image without generating one? Send in a wide image directly, or open the slicer: https://youmind.com/a/dlJ7lVfZ9ZVqFi