Free tool

AI video prompt builder with real cuts

Describe each cut: its shot size, camera move and the one thing that happens. Get a timed, ready-to-paste prompt for Kling or Wan, the way a director hands a shot list to an editor.

Who is in itWhere and when

Cuts · 7s

Cut 1

Cut 2

Cut 3

Cut 4

Cut 5

Sound

I have an end frame

Your prompt

7-second vertical cinematic social reel, photoreal, 5 cuts. The woman from the start frame; keep the same face, hair and outfit exactly throughout. A small Paris apartment on a rainy afternoon. The video opens on the start frame and ends exactly on the end frame.
Cut 1 (0.0-1.4s): Medium shot, slow push-in: she stares at her dark phone, waiting; one eyebrow slowly rises.
Cut 2 (1.4-2.4s): Extreme close-up, locked-off: her fingernails tap impatiently on the table beside the phone.
Cut 3 (2.4-3.8s): Wide shot, handheld: she opens the window and releases a white pigeon into the rain.
Cut 4 (3.8-5.0s): Wide shot, whip-pan: the pigeon flies fast over grey rooftops.
Cut 5 (5.0-7.0s): Medium shot, locked-off: she sips her tea, smug, while the phone on the table lights up again and again.
Sound: rain on the window, nail taps, wingbeats, a phone buzzing. No dialogue. No music. No text, subtitles or captions.

Use it with an image-to-video model and a start frame you approved first. With an end frame (Kling supports one), the last cut is guaranteed to land on it.

Create your influencer

Free to create, no card needed. The first 1,000 get 100 extra credits.

Tips

How to write cuts a video model follows

One change per cut

Each cut does one thing: she looks up, the bird takes off, the phone lights up. Ask for three things in one cut and the model does one, or a blur of all three.

Describe what changes

The start frame already fixes the face, outfit, room and light. The prompt only needs what moves and how the camera moves.

Vary the shot sizes

Cutting from a wide shot to an extreme close-up of hands to a reaction makes a reel feel edited. Staying in one medium shot feels like a webcam.

Draw the payoff as an end frame

Models struggle to change something mid-shot, like a screen switching on. Draw the final moment and pass it as the end frame so the reel has to land there.

Write the sound, skip the music

A sound line (rain, footsteps, a door) makes clips feel real. Some models add music anyway; replace the audio when you need your own.

Keep scripts out

Never paste captions or a script into a video prompt unless someone must say it: models tend to print it on screen as subtitles.

FAQ

Questions, answered

What is a cut list prompt?

A video prompt written like a director's shot list: a header with the length and what stays the same, one timestamped line per cut with its shot size, camera move and one action, and a sound line. Models like Kling and Wan follow it closely.

Which models does it work with?

It writes prompts in the format Kling 3 and Kling 4 follow ("Cut 1 (0.0-1.4s): …") and the timestamped format Alibaba's guide recommends for Wan ("Shot 1 [0-1.4 s] …"). Other image-to-video models generally follow the same structure.

How long should each cut be?

For short-form reels, about 1 to 2 seconds a cut, and a final shot held a little longer so the payoff lands. A single long shot of up to 30 seconds also works if it is described well.

Is the prompt builder free?

Yes. It runs in your browser, needs no sign-up and nothing you type is sent anywhere. Copy the prompt and use it with your own video model.

What does a reel made this way cost?

In our tests a 10 to 15 second reel cost about $1.50 to $3 on Kling or Wan. Our free AI video cost calculator works it out for your settings.

AI video prompt builder with real cuts

Describe each cut: its shot size, camera move and the one thing that happens. Get a timed, ready-to-paste prompt for Kling or Wan, the way a director hands a shot list to an editor.

Who is in itWhere and when

Cuts · 7s

Cut 1

Cut 2

Cut 3

Cut 4

Cut 5

Sound

I have an end frame

Your prompt

7-second vertical cinematic social reel, photoreal, 5 cuts. The woman from the start frame; keep the same face, hair and outfit exactly throughout. A small Paris apartment on a rainy afternoon. The video opens on the start frame and ends exactly on the end frame.
Cut 1 (0.0-1.4s): Medium shot, slow push-in: she stares at her dark phone, waiting; one eyebrow slowly rises.
Cut 2 (1.4-2.4s): Extreme close-up, locked-off: her fingernails tap impatiently on the table beside the phone.
Cut 3 (2.4-3.8s): Wide shot, handheld: she opens the window and releases a white pigeon into the rain.
Cut 4 (3.8-5.0s): Wide shot, whip-pan: the pigeon flies fast over grey rooftops.
Cut 5 (5.0-7.0s): Medium shot, locked-off: she sips her tea, smug, while the phone on the table lights up again and again.
Sound: rain on the window, nail taps, wingbeats, a phone buzzing. No dialogue. No music. No text, subtitles or captions.

Use it with an image-to-video model and a start frame you approved first. With an end frame (Kling supports one), the last cut is guaranteed to land on it.

Create your influencer

Free to create, no card needed. The first 1,000 get 100 extra credits.

AI video prompt builder with real cuts