A repeatable 6-part formula for writing prompts that come back looking intentional instead of generic — with weak-vs-strong examples for cinematic, product, and anime styles.
Aug 19, 2026 · 7 min read · AI Image Generator

Most disappointing AI images aren't a generator problem — they're a prompt problem. A vague request forces the model to guess at style, lighting, and composition, and it usually guesses generic. Here's the exact 6-part structure that removes the guesswork.
Subject
What's actually in the frame — a person, object, scene, or character.
"a lone astronaut"
Style
The rendering approach — cinematic, 3D, anime, realistic, or product.
"cinematic lighting"
Setting
Where the subject is, and what's around it.
"standing on a red desert dune"
Lighting
The light source and mood it creates.
"golden hour backlight"
Composition
Framing and camera angle — close-up, wide shot, low angle.
"wide shot, low angle"
Detail level
How much texture and finish to render.
"sharp focus, high detail"
Weak
"astronaut on mars"
Strong
"A lone astronaut standing on a red desert dune, cinematic lighting, golden hour backlight, wide shot, low angle, sharp focus, high detail."
Why it works: The weak version leaves every creative decision to chance. The strong version tells the generator exactly what style, light, and framing to commit to — which is why it comes back looking intentional instead of generic.
Weak
"cool product photo"
Strong
"A skincare bottle on a clean marble surface, product photography style, soft studio lighting, straight-on composition, sharp focus, high detail, clean background."
Why it works: "Cool" means nothing to a generator. Naming the actual style (product photography), the light setup, and the composition removes the guesswork entirely.
Weak
"anime girl"
Strong
"A samurai warrior standing in a bamboo forest, anime style, clean line work, dramatic rim lighting, close-up composition, sharp focus, high detail."
Why it works: Naming a specific subject and setting gives the generator something to actually compose around, instead of defaulting to the most generic version of the request.
Subject and style carry most of the weight — if you only have time for two parts, start there. Setting, lighting, composition, and detail level are what separate a good image from one that looks deliberately art-directed, so add them once you're generating regularly.
Pick a style — cinematic, 3D, anime, realistic, or product — and write one prompt using all six parts. Then generate it in Zyvo's AI image generator.
Open the AI Image Generator →Real preset screenshots from the generator, with the exact opening lines used in each.
18.08.2026
How to choose the best AI image generator for social media in 2026.
11.02.2026
Anime, 3D, or realistic — use only the best AI image styles and learn to control them.
12.03.2026
Real output across Zyvo's cinematic, 3D, and realistic styles, and what each one is actually good at.
19.08.2026