How to write an AI image prompt that comes out right

Image AI is very literal. It gives you exactly what you describe, which is the problem, because most of us describe pictures the way we describe dreams: one word and a vibe.

A prompt that works has a few parts. You do not need all of them every time, but the more you include, the less the model has to guess.

The subject

What is in the picture. “A candle” is a subject. “A hand-thrown ceramic candle holder with a lit beeswax candle” is a subject you can use.

The details

The things that matter to you. Shape, material, color, how many items, what is in the background. This is where a prompt is won or lost.

The style

Realistic photo, illustration, flat design, watercolor, 3D render. Skip it and you get whatever the model feels like that day.

The lighting and mood

“Soft morning light through a window”, “warm and cozy”, “clean studio lighting on white”. These change the result more than people expect.

The format

Square for Instagram, wide for banners, tall for product shots. Say it in words.

Here is the same idea twice.

Weak: “a candle for my shop listing”

Strong: “A hand-thrown ceramic candle holder in cream, holding a lit beeswax candle, on a rustic wooden table, soft window light from the left, cozy warm mood, photorealistic product photo, square format, plain blurred background”

The second one is a prompt you can run. It will still not be perfect, and that is normal. Check the details that matter: text in the image, hands, logos. Those are the parts image AI gets wrong most often. If something is off, change one part and run it again. The prompt is a draft, not a contract.

Writing these from scratch every time gets old, so I made a tool for it. The AI image prompt generator is an AI tool: describe the picture you want in plain words, and it writes out a detailed prompt ready for Midjourney, Flux, or any image tool. One credit per run. It does the same job as the example above: takes your half-formed idea and fills in the parts you forgot.