AI Image Prompt Builder

How to Build a Prompt

1. Describe your subject. Start with what you actually want to see — the person, object, or scene at the centre of the image.
2. Add the details. Work through the options to set things like style, lighting, mood, and quality. Each choice appends another descriptive phrase to the prompt.
3. Review the assembled prompt. The full text builds up as you go, so you can see exactly what the generator will receive.
4. Copy it into your generator. Paste the result into Midjourney, Stable Diffusion, DALL·E, or whichever tool you're using.

What This Tool Does

Most disappointing AI images come from prompts that are too vague. "A cat" leaves every decision to the model; "a tabby cat asleep on a windowsill, soft afternoon light, shallow depth of field, photographic" leaves far less to chance. This builder walks you through the details worth specifying and assembles them into a properly structured prompt, so you're not trying to remember what to include each time.

It works with any image generator, and everything runs entirely inside your browser — no account, no API key, no images processed here.

What Makes a Prompt Work

A strong prompt usually covers four things: the subject (what's in the frame), the style (photographic, illustrated, 3D render, and so on), the lighting and mood (golden hour, moody, high key), and technical qualities (depth of field, camera angle, level of detail). Miss one and the model fills the gap with whatever is most common in its training data, which is why vague prompts tend to produce generic, oddly similar results.

Order matters too. Most generators weight words near the beginning of a prompt more heavily, so lead with what's essential and leave refinements for the end. Beyond that, treat the first result as a starting point rather than a verdict — change one element at a time and regenerate. Changing five things at once tells you nothing about which one helped. And bear in mind that each generator has its own syntax and quirks: Midjourney takes parameters that Stable Diffusion doesn't understand, so a prompt that shines in one may need adjusting for another.

Frequently Asked Questions