Related: the full prompt library, product ad prompts, and video ad prompts.
Related: the full prompt library, product ad prompts, and video ad prompts.
Every prompt on this page was made with ChatGPT's image model, GPT Image, and sits next to the image it produced. They cover product shots, posters, static ads, infographics and UGC-style photos. Open a card to copy the full prompt, paste it into ChatGPT, attach your own product photo, and ask it to swap your product in.
ChatGPT's image model is good at what ads need: readable headline text, layouts with several elements, and long, detailed instructions. That suits it to posters, infographics and static ads with copy on them. For pure product photography, many prompts in our main gallery were made with Nano Banana, so try both on the same brief.
Describe the ad, not just the picture: the product, setting, light, camera angle, any headline text in quotes, where that text sits, and the aspect ratio. Then refine with short follow-up instructions instead of starting over.
Yes. GPT Image is one of the better models for text inside images, so headlines and short labels usually come out readable. Proofread before you run the ad.
Yes. Copy and adapt any prompt for your own ads. Open a card to see the creator who shared it.