Prompt anatomy: say what you actually mean.
The gap between a generic image and a striking one is almost always the prompt. Here is the structure professionals use — and a pattern you can reuse for anything.
01The five levers
A strong prompt usually names five things: the subject (what), the medium (photo, oil painting, 3D render), the style (artist, era, aesthetic), the lighting (golden hour, studio, neon), and the composition (close-up, wide shot, top-down). Name them and the model stops guessing.
A reusable prompt skeleton[subject, described concretely], [medium], [style or reference], [lighting], [composition / camera], [mood]
Example: a weathered fisherman mending a net, documentary photograph, muted color, soft overcast light, tight close-up on the hands, quiet and contemplative
02Concrete beats clever
"Beautiful" and "amazing" do almost nothing — the model already aims for those. Specific nouns and verbs do the work: not "a nice city," but "a rain-slicked Tokyo alley at night, steam rising from a ramen stall." Describe what a photographer would actually frame.
Word order matters — Midjourney weights earlier words more. Lead with the subject and the one or two things you care most about, then add supporting detail.
Frequently asked
Midjourney — your questions, answered
How do I write a good Midjourney prompt?
Name five things: subject, medium, style, lighting, and composition. Lead with the subject, use concrete nouns and verbs, and add supporting detail after.
Does word order matter in Midjourney prompts?
Yes — earlier words carry more weight. Put the subject and your most important details first.
Why do my images look generic?
Usually the prompt is too vague. Swap "beautiful" and "amazing" for specific, concrete description of what should be in the frame and how it is lit.
Is there a reusable prompt formula?
Yes: [subject], [medium], [style], [lighting], [composition], [mood]. Fill each slot concretely and refine from the grid.
Keep going
← Back to all Midjourney lessons