Skip to main content
Free tool · No account

AI prompt enhancer for images and video

Paste a rough idea and get it back as a fuller prompt for image, video or music. The six-point checklist it works from is printed below.

Free · No signup · Up to 2000 characters

The enhancer

Rough in, specific out

Pick what you are making, because the three want different things: an image prompt is a description, a video prompt is a description plus a movement, and a music prompt is closer to a list of tags than a sentence.

What are you making?

The three are written differently, so pick before you paste.

0 / 2000 characters

The checklist

What a prompt has to answer

Most weak prompts are not badly written, they are underspecified — the engine is left to invent the half you did not mention, and it invents something different every time. These are the 6 decisions worth making yourself.

Image

A still frame is a description. Everything you leave out gets invented.

Subject
Who or what, and in what state. “A woman” is a placeholder; “a woman shielding her eyes from low sun” is a picture.
Setting
Where it is, and roughly when. A time of day does more work than a list of adjectives.
Light
Direction and quality. Backlit, overcast, single hard source — this is what separates a photograph from a render.
Framing
How close, and from where. Close-up, wide, low angle, over the shoulder.
Style
A medium and a treatment: 35mm film, flat vector, watercolour. Name one, not five.
What to leave out
The thing the engine keeps adding that you do not want — text on the image, extra people, a logo.

Video

A clip is a description plus what changes over its length — including the camera.

Subject
Who or what is on screen, described the way you would for a still.
Action
What actually changes across the clip. A clip with no action is a photograph with a runtime.
Camera
Static, slow push, handheld, orbit. If you do not say, you get a slow drift you did not ask for.
Pace
Whether it is one continuous movement or a held frame. Fast cuts are an edit, not a prompt.
Setting and light
Where it happens and how it is lit, same as for an image.
Length
How long the shot runs. A four-second beat and a ten-second one are different shots, not the same shot cut down.

Music

A track prompt is closer to a tag list than a sentence. Genre first, always.

Genre
The first tag and the most load-bearing one. Everything else is read relative to it.
Mood
Two or three words. Driving, melancholy, warm — not a paragraph of feeling.
Instruments
The two or three that define it. Naming twelve gets you a mush of all of them.
Tempo and energy
Slow, mid, driving, or an actual BPM if you know it.
Vocal
Instrumental, or a voice and its character. Silence here usually gets you a vocal you did not want.
Structure
Where it goes: builds to a chorus, stays flat as a bed, ends cold.

Work down a column and you will write a better prompt than the enhancer will, because you know what you meant. The tool is for when you want it done in one step instead of six.

Before and after

What the difference looks like

Three rough prompts, what the enhancer returned for each, and the image that returned prompt made at 2K.

  • A red fox with fluffy fur sitting in a snowy forest in morning sun

    Enhanced prompt → image · 2K

    What people type

    “fox sitting in snow forest, morning sun, close up, fluffy fur, photo realistic 4k”

    What the enhancer returned

    “A close up photograph of a fox with fluffy fur sitting in a snowy forest, illuminated by morning sun.”

  • A steaming coffee cup by a rainy window with blurred city lights outside

    Enhanced prompt → image · 2K

    What people type

    “coffee cup rainy window cozy vibe, steam, city lights bokeh outside, warm tones, film look”

    What the enhancer returned

    “A coffee cup with steam rising from it sits near a rainy window, creating a cozy vibe. Outside the glass, city lights blur into bokeh. The scene features warm tones and a film look.”

  • Glowing blue and pink jellyfish with long tentacles in the dark ocean

    Enhanced prompt → image · 2K

    What people type

    “jellyfish glowing blue and pink deep dark ocean, long tentacles, macro, dreamy”

    What the enhancer returned

    “A macro shot of glowing blue and pink jellyfish in the deep dark ocean, featuring long tentacles in a dreamy style.”

Questions

The ones worth answering

What does the enhancer actually change?

It keeps your intent and fills in what you left unsaid: the setting, the light, the camera. It does not add subjects you did not mention, and if it cannot improve your prompt it says so.

Is it free?

Yes, with no account and no email. Each rewrite costs us something, so there is a shared daily limit; when a day runs out, the page says so and the checklist above still works.

Which model does it use?

We do not name it: the engine behind a feature changes when a better one appears. What stays fixed is the behaviour: your intent is kept, and the output fits the length the target engine accepts.

Do you keep the prompts I type?

Your prompt is sent to be rewritten and returned, and it is not tied to an account because there is none. Do not paste anything confidential into any free web tool, ours included.

What if my prompt is already long?

It is tightened rather than padded. Engines ignore text past their input limit, which is worse than a short prompt; the upper bound here is 2000 characters.

Can I use the result in another tool?

Yes. The output is plain text and nothing about it is specific to us. It is written against how image, video and music engines generally read a prompt, so it transfers — and you are welcome to take it elsewhere.