How to write an image prompt for a meditation cover or thumbnail
A good image prompt names one clear subject, where it is, how it should feel, the light, and the style, then says what to leave out. That's the order this tool writes it in, so the most important words come first.
Start with the scene, not the session
"A quiet lake at dawn" gives an image generator something to draw. "A relaxing meditation for stress" doesn't. Describe what you'd see, then let the mood and light carry the feeling.
Pick the shape first
A YouTube thumbnail is wide (16:9), cover art is square (1:1) and a Story or Reel is tall (9:16). Getting the shape right up front saves cropping later, and leaving empty space on one side gives your title somewhere to sit.
Leave out text, logos and faces
Image generators often garble lettering, so it's usually cleaner to add your title afterwards in Canva or your editor. Leaving out recognisable people and brands also keeps your artwork original and avoids likeness and trademark problems.
Why the prompt changes for each generator
Midjourney reads settings at the end of a prompt, such as "--ar 16:9" for the shape and "--no" for things to leave out. Leonardo has a separate negative prompt field. ChatGPT works best with plain sentences. The creative idea stays the same; only the format changes.
Is my description sent anywhere?
No. The prompt is built in your browser from your choices. Nothing is sent to Jaapi or to an AI service until you paste it into the image generator you use.