How it works
You start with one sentence.
You finish with a full kit.
Let's follow the same video you saw on the home page — stop by stop, with what you type in along the way and what comes back.
“A video about why your website is slow. For business owners, 40 seconds, serious tone.”
-
1
You say what it is about
2 minFive questions, one at a time. What it is about, who it is for, where it will run, how long it is, what the tone should be. No blank editor, no „start writing something".
what you type in- topic: why your website is slow
- audience: small business owners
- length: 40 seconds · tone: serious
From this the system already knows that three or four scenes will fit, and that the tone cannot be playful.
-
2
It writes the script
1 minOne press, and the AI writes from what YOU gave it — not in general terms, but for your topic and your running time. Broken into scenes, timed to the second.
what you have by now- 0–6 s · „Three seconds. That is all you get."
- 6–22 s · the three most common causes
- 22–40 s · what to do today
Each scene comes with the spoken text in your own language and the shot description in English, for the generator.
-
3
You pick who and what is on screen
5 minYour character's looks, age and clothing — by clicking, with color swatches and drawn samples. Then the location, the time of day and the color. The character builder even lets you switch on small lifelike flaws, so the face doesn't come out plastic.
what you have by now- man, 40s, short beard, grey shirt
- tired eyes, slight asymmetry
- small office, late afternoon light
From here on every later step knows your character — from the shot list to the image prompt.
-
4
The image prompt comes together
8 minLens, camera angle, lighting, film look, aspect ratio. The English text builds live on the right, with a plain-language version underneath — and the blocklist goes in too, saying what the machine should NOT do.
what you have by now- cinematic medium shot, desk, 35mm lens
- f/1.8, shallow depth, teal-orange, 16:9
- slow push-in
This is what you paste into your generator. The camera move comes separately, for the video step.
-
5
Then the music and the sound
3 minWe suggest the musical style from your tone, and take the cut points from the scene breaks in your script — you don't have to measure them out. The voiceover comes with tone, pace and pauses.
what you have by now- tense minimal underscore, 96 BPM
- cut points: 6 s · 22 s
- voiceover: calm, low register
If you change the running time later, the music prompt flags itself as stale — you never carry on with the old one.
-
6
The kit is done
2 minTimed SRT captions, an edit plan, a title, a cover-image prompt and several post versions. Each one a separate, copyable piece — not one big block you have to dig through.
what you have by now- captions.srt · edit plan
- title + cover-image prompt
- 3 post drafts
From here your own generator and editing software take over. The plan is done.
What happens on its own meanwhile
Three things run in the background throughout. You never have to think about any of them — but it helps to know they are there.
Including as you type, on a short delay. There is no „Save" button to forget, and nothing is lost if a tab closes by accident.
Your character description feeds the shot list; the scene breaks in your script feed the music cut points. Nothing has to be typed twice.
Change the running time from 38 to 45 seconds and the music prompt is marked stale straight away. So you never carry on with an old figure without noticing.
Your own generator takes it from here
We're not competing with them — we work for them. The prompts come out in English because that is what they read most precisely, and they paste into any of them.
- Midjourney
- GPT Image 2
- Nano Banana Pro 2
- FLUX.2
- Runway
- Kling
- Veo
- Higgsfield
- ElevenLabs
- Suno
The names above belong to their respective owners; we're not affiliated with any of them.
Common questions
How long does all this take?
Give your first video about an hour, since you're learning the interface at the same time. By the second, the plan takes about twenty minutes — then comes the generating, which takes however long your own generator takes.
Do I have to do all fifteen steps?
No. You pick the project type at the start: if you only need a prompt for a single image, you get three steps, not fifteen. The full run is for a complete movie.
What if I stop halfway?
Nothing. Every field saves itself, including as you type. When you come back, the system shows you where you left off and you carry on from there.
Can I change something later?
Yes, whenever you like. Change an earlier answer and the system marks which finished texts have gone stale because of it — so an old running time or a tone you have since changed never survives into your video.
Where do I paste the finished prompt?
Into whichever generator you use: Midjourney, FLUX, Nano Banana, Runway, Kling, Veo, Higgsfield — the prompts come out in English because that is what they read best. Suno or similar for the music, ElevenLabs for the voice.
Run it yourself
The same six stops, with your own topic. Everything is yours for two weeks, no card.