- Free
- No sign-up
- 20 prompts a day
AI Video Prompt Generator
Describe a scene in a few words. Get a shot-by-shot video prompt with camera moves, timing, lighting and sound, then generate the clip in one click.
0/500
Need an idea? Try one
Your shot-by-shot prompt appears here
The tool keeps what you wrote and fills in:
- 1Overall look
- 2Timed shots
- 3Shot size and camera
- 4Subject action
- 5Light and color
- 6Sound and text
Why video prompts need camera language
A video model has to decide far more than an image model: how the camera moves, how close it is, what happens first and how long each moment lasts. Leave those out and the model guesses, which is why short prompts give jittery, aimless clips.
This tool keeps everything you asked for and writes it the way a director would brief a crew. It splits your clip into timed shots, gives each one a shot size and a camera move, keeps the subject consistent from shot to shot, and adds light, color and a line of sound. You get one paragraph you can read, edit and paste into any text-to-video model.
How to write an AI video prompt in seconds
- 1
Describe the scene
Write what should happen in a sentence or two, in English or Chinese. Add any words that must appear on screen.
- 2
Pick style, ratio and length
Choose cinematic, anime, documentary, commercial or 3D, the aspect ratio for where the video goes, and 5, 10 or 15 seconds.
- 3
Generate, edit and use it
Change any shot you like, copy the prompt, or send it to Seedance 2.0 with the ratio and length already set.
What a good video prompt contains
Every prompt this tool writes covers these parts. Use the same checklist when you write video prompts yourself.
1Look
Cinematic live action, 16:9, 10 seconds
Medium, orientation and total length set the frame for everything else.
2Shots
0-3s: wide, low angle, tracking shot
Time ranges in order. 1-2 shots for 5s, 2-3 for 10s, 3-4 for 15s.
3Camera
slow push-in, pan, orbit, crane, handheld
One clear move per shot beats a vague "dynamic camera".
4Action
the dog leaps toward the drifting red balloon
What the subject does in each shot, in plain verbs.
5Continuity
same tan shiba, red collar, in every shot
Repeat the subject's look so it doesn't change between shots.
6Light and sound
neon reflections, film grain; rain and distant traffic
Mood comes from light, color and one line of sound.
Styles you can choose
Cinematic
- Film-like live action with deliberate camera moves and dramatic light. Good for stories, travel and mood pieces.
Anime
- 2D animation with cel shading and expressive motion, for character scenes and music visuals.
Documentary
- Natural, observational footage with steady or handheld shots, for nature, city life and behind the scenes.
Commercial
- Polished product and brand spots with clean studio or lifestyle shots and an end card when you ask for one.
3D
- Stylized 3D animation for mascots, toys, explainers and product concepts.
Auto
- Lets the tool pick the look that fits your idea. A good start when you're not sure.
Tips for better AI videos
One idea per clip
A 5-10 second clip holds one moment well. Split a longer story into several prompts.
Name the move you want
"Slow push-in" or "orbit around the bottle" works far better than "cool camera".
Keep the subject consistent
Describe clothing and colors once and keep them. Changing them mid-prompt makes the model swap the subject.
Match the platform
9:16 for Reels, TikTok and Shorts, 16:9 for YouTube and web, 1:1 for feeds.
Quote on-screen text
Put the exact words in quotes. The prompt places them and asks the model to add nothing else.
Edit one shot at a time
If one shot comes out wrong, change only that time range and keep the rest.
Frequently asked questions
Is this AI video prompt generator free?
Yes. There is no sign-up and no watermark. You get 20 successful prompts every 24 hours, and failed attempts don't count.
Which video models do the prompts work with?
The prompts are plain descriptive language with timed shots, so they work in Seedance, Veo, Kling, Wan and other text-to-video models.
Why does the prompt have time ranges?
Video models follow a sequence better when each part has a time slot. The tool splits 5, 10 and 15 second clips into 1-4 shots so the action fits the length.
Can I write my idea in Chinese?
Yes. Write in English or Chinese. The prompt comes out in the language of this page, and any text you want on screen stays exactly as you wrote it.
Can I generate the video right after?
Yes. Select the generate button and the prompt, aspect ratio and length are filled in on the Seedance 2.0 model page.
Do you store my ideas?
No. Your idea is sent to the language model only to write this prompt and is not saved.
Can I use this in my own app?
Yes. Create a HiAPI API key, call a text model through the OpenAI-compatible /v1/chat/completions endpoint to write prompts, and generate videos through the same API.
Generate videos with the HiAPI API
One API key for Seedance, Veo, Kling, Wan and more. Pay as you go, no subscription.