At the start of your ChatGPT conversation, paste in this one-time prompt (remember you can adjust the prompt to your own needs and tools, and I highly recommend you do so!)

Locate the #Task section and write a simple line of what you want in the image. Example: A close-up of Merynia sitting on her couch while drinking coffee. Merynia has flowing blue hair, pale skin and wears glasses.
The system prompt returned this:
Captured with a Canon EOS R5 and an RF 85mm f/1.2 lens for crisp detail, shallow depth of field, and cinematic realism, the image shows a close-up of Merynia sitting comfortably on her sleek modern couch, sipping coffee with a relaxed yet confident expression. Her flowing blue hair frames her pale skin elegantly, her stylish glasses catching the soft morning light filtering through tall windows. The environment is bright and minimalist, with subtle props like a silver laptop, an open notebook, and a smartphone on the side table—symbols of the digital entrepreneur lifestyle. Warm natural lighting mixes with soft shadows to create a cozy yet powerful atmosphere, suggesting both calm and control. The mood embodies freedom, success, and transformation through AI and online income: the simple joy of working on your own terms, escaping the 9-5, and thriving in a modern, balanced life. Perfect for vertical reels and widescreen thumbnails, the image feels aspirational but attainable, inviting viewers to imagine themselves living this same reality of digital independence.

Now, whenever you want a new image, you don’t need to rewrite all that again. Just write a simple line explaining what you want. Example: Now a close-up of Merynia in a tennis outfit, getting ready to leave for her match

That’s it. ChatGPT will remember the camera, lens, vibe, goals, style, brand feel… all the rules. You only write what changes.