How to create great AI images: a beginner's prompt guide
A beginner's guide to AI image generators (ChatGPT, Gemini, Grok, Midjourney, Canva): the prompt structure, styles, fixes, and usage rights basics.
Pick a tool
Chatbots like ChatGPT, Gemini and Grok can generate images directly; Canva and Adobe Express have built-in generators; Midjourney is popular with designers. Free limits change often, so check each tool.
Use the subject-scene-style-light-format structure
A golden retriever wearing a yellow raincoat (subject), sitting at a bus stop in light rain (scene), children's book watercolor illustration (style), soft morning light (lighting), wide 16:9 format (format).Iterate in small steps
Change one thing at a time: "same image, but the coat is red" or "closer crop on the face". Big rewrites lose what you liked.
Useful style words
- Photo: "35mm photo, shallow depth of field, natural light"
- Graphics: "flat vector illustration, limited palette, white background"
- Product: "studio product shot on seamless grey background, soft shadows"
Common fixes
- Garbled text in images: keep text very short and put it in quotes, or add text later in Canva.
- Weird hands: regenerate or crop.
- Generic look: add specific details: materials, era, location, mood.
Rights and honesty
Read each tool's terms for commercial use, don't imitate living artists by name or create images of real people without consent, and label AI images where it matters (news, reviews, real estate).
Quick tips on this topic
⚡ The image prompt formula
Subject, scene, style, lighting, format. Fill all five for images that look intentional.
⚡ Change one thing at a time
Loved an image except one detail? Ask for that change only.
⚡ Add text later
AI still garbles text in images. Keep words short, or add them in Canva after.
Swipe 50+ quick AI tips →One tip per screen, 10 seconds each