Free Script to AI Video Prompts Converter
Paste a script and this free tool splits it into scenes, then writes full, ready-to-paste video prompts for the first two. You’ll see how many scenes your script has, so you know exactly what it takes to make the whole video.
It works with any format: proper scene headings, a voiceover script, a list of lines, or plain paragraphs. Pick your video model and the prompts come out in that model’s style.
How to turn a script into AI video prompts
One scene, one prompt
Video models make one shot or one short sequence at a time. Each scene in your script becomes its own prompt, timed to what is said in it.
Show what the words mean
If the line is "I was broke", don’t show a person thinking. Show the empty wallet, the overdue bill. Good scene prompts turn words into pictures.
Keep the character identical
Video models forget your character between clips. Describe them with the same exact details in every scene: age, hair, clothes, one distinguishing detail.
Fast cuts in the hook
The first three seconds decide if people stay. Give the opening scene two or three quick, striking shots before the story settles.
Time scenes to the voice
About 150 spoken words is a minute of video. A scene with 25 words of voiceover needs about 10 seconds of footage, so pick clip lengths to match.
Direct the audio per scene
If your model makes sound, each scene prompt should say what we hear: the line being spoken, ambience, or "no dialogue" for B-roll.
Script to video prompt examples
Copy any of these and paste it straight in, or use it as a starting point.
Voiceover line → scene
Script: VO: "Month two is where most side hustles die."
Prompt: Shot 1 (0-1.5s): Extreme close-up, static. A hand tears a page off a wall calendar, revealing the next month. Shot 2 (1.5-4s): Medium shot, slow push-in. A 27-year-old man in a grey hoodie sits at a cluttered desk, staring at a laptop showing a flat sales chart. Warm desk lamp against cool laptop glow. Sound: paper tearing, then quiet room tone.
Dialogue scene
Script: MAYA: "You kept it?" / LEO: "Every single one."
Prompt: Single continuous shot, no scene cuts. Medium two-shot at eye level. A woman in a yellow raincoat holds up a shoebox of old letters in a dim attic and says, "You kept it?" The man beside her, in a flannel shirt, smiles and says, "Every single one." Dusty afternoon light through a round window. No music.
B-roll for a product line
Script: "It charges your phone four times on one trip."
Prompt: Close-up, slow orbit left. A matte black power bank lies on a wooden café table beside a passport and a paper coffee cup, a phone charging from it, the screen lighting up as the battery fills. Soft daylight, warm wood tones. Sound: café murmur, a short charging chime. No dialogue.
Opening hook scene
Script: "Your expensive serum is mostly water."
Prompt: Shot 1 (0-1s): Macro, static. A drop of serum falls into a glass of water and disappears. Shot 2 (1-3s): Close-up, handheld. A woman turns a small serum bottle in her fingers, squinting at the ingredient label. Bright bathroom light, white tiles. Sound: a soft drip, then a quiet scoff. No music.
Closing scene
Script: "Keep going. There’s room for you tomorrow."
Prompt: Wide shot, slow crane up. A young woman stands on a rooftop at sunrise, laptop bag on her shoulder, looking out over a waking city. Soft pink and gold light, long shadows across the rooftop gravel. Hopeful, still. Sound: distant traffic starting up, a light wind, no dialogue.
Common mistakes when turning scripts into video
One prompt for the whole script. A model can’t fit a minute of story into one clip. Split it into scenes first.
A new face in every scene. If each prompt describes the character differently, you get a different person every clip.
Visuals that ignore the words. Random pretty footage under a voiceover feels like a slideshow. Each scene should show what is being said at that moment.
Clips that don’t match the voice. A 3-second line in a 10-second clip leaves dead air. Time each scene to its words.
Cutaways inside a talking clip. If someone speaks on camera, cutting away mid-sentence breaks lip-sync. Keep talking clips to one take and put B-roll in separate clips.
One prompt is one shot. A video is 6–10.
This free tool writes one great prompt. ScriptSea makes the whole video plan: you give it one idea, it writes a scroll-stopping hook and a full script, then a ready prompt for every scene, with the same character in every shot. It works with every major video model.
This free tool
- One prompt at a time
- First 2 scenes of a script
- You write the script yourself
ScriptSea
- Idea → hook → full script
- A ready prompt for every scene
- Same character across all scenes
- Any model: Seedance, Veo, Omni Flash, Kling and more
3 free scripts. No card.
Script to Video Prompts FAQ
How do I convert a script into AI video prompts?
Paste the script, pick your video model and press Generate. The tool splits it into scenes and writes a ready-to-paste prompt for each of the first two.
Why are only two scenes unlocked?
The free tool writes the first two scenes so you can see the quality. ScriptSea unlocks every scene, keeps the same character in each one, and can write the script for you too, with 3 free scripts and no card.
What script formats work?
Anything: screenplay format with scene headings, a voiceover script, dialogue lines, or plain paragraphs.
How many scenes does a short video need?
Usually 6 to 10 for a 30 to 60 second video. Each scene is one clip of about 3 to 10 seconds.
Will the character look the same in every scene?
The prompts describe the character with the same details each time. For perfect matching, also use a reference image in your video tool.
Does it work for other languages?
Yes. The prompts come back in English for the best results, and any dialogue stays in your script’s language.