Text to video
AI text to video, starting with your idea
Describe the scene you want to see and generate a short video from text. Explore an opening shot, a social video idea or a visual mood before you have a finished image to work from.
Open the generator to choose a mode, model and settings. Review the credit cost before generating.
Prompt structure
Subject + action + scene + camera
A paper boat drifts across a shallow puddle after the rain. Warm evening light reflects in the water. The camera stays low and slowly follows the boat. One quiet, continuous shot with a realistic look.
How to turn a description into a video
Give the model a coherent shot to work with, then refine the part that matters most to your idea.
- 01
Write one clear scene
Select Text to video in the generator. Describe the main subject, its action and where the scene takes place. Add camera movement, lighting and style only when they serve that shot.
- 02
Set the shape and length
Pick a model and choose from its supported duration, aspect ratio and resolution options. Match the frame to your intended placement and review the displayed credit cost.
- 03
Review and refine the shot
Watch the generated clip for subject clarity, timing and unwanted details. Change one part of the prompt, generate another version if needed, and download the clip you choose.
Write for a short clip
Build a shot before you build a story
A short generation is easier to direct with a single visual idea. For a longer sequence, create separate clips and assemble them in your editing workflow.
Subject: a cyclist in a yellow raincoat. Action: riding slowly along an empty coastal road. Scene: overcast morning, sea mist and wet pavement. Camera: smooth side-tracking shot, medium-wide framing. Style: natural colors, realistic light, one continuous take.
Prompt example only. Use concrete actions and avoid conflicting camera instructions in the same shot.
Choose the right starting point
Text to video is useful for exploring a scene that does not need to match a specific source image.
Use text for new visual ideas
Atmosphere shots, visual concepts and simple actions can begin with a description. Tell the model what should be visible, rather than relying only on abstract words such as “premium” or “viral”.
Use an image for a specific product
If a real product, package or character must guide the appearance, Image to video gives the model a visual reference. Text alone cannot reliably specify every visual detail.
Finish exact copy after generation
Review generated text, logos and fine details carefully. Add exact titles, prices or captions in your editing workflow, and check audio support for the model you choose.
Text to video questions
Do I need an image to start?
No image is required for Text to video. Start with a scene description. If you already have an image whose appearance should guide the result, choose Image to video instead.
What should I include in a prompt?
Describe the subject, one main action and the environment, then add camera movement and lighting. Keep instructions consistent with the clip length and make one change at a time when refining a result.
Can it create a complete edited ad?
The generator creates video clips. A finished ad may also need several shots, verified product claims, exact copy, captions, music or voiceover. Plan those elements and complete any remaining assembly in your editing workflow.
Start with one scene you can describe
Write the subject, the action and the camera. Then explore the first version in the generator.
Create a video from text