What is text-to-video AI?
A text-to-video model creates a short moving scene from a written prompt. Your description supplies the subject, environment, action and visual direction. It is useful when you want to explore an idea without preparing a starting image.
Build a prompt in four parts
- Subject: name the main thing the viewer should notice.
- Environment: describe the location, time of day and visual atmosphere.
- Action: explain what changes during the shot.
- Camera: choose a simple framing or movement, such as a fixed view or a gentle forward move.
Try a scene with a clear direction
A small glass house sits in a misty pine forest at dawn. Warm light turns on inside as fog moves between the trees. The camera glides slowly forward. Calm, realistic lighting, one continuous shot.
This is a prompt draft, not a tested performance claim. Change one element at a time when comparing results. Too many subjects, rapid scene changes or conflicting camera instructions can make a result harder to judge.
Choose settings for the clip you need
The creator lists available models and their supported combinations. Review duration, resolution, aspect ratio and any audio control before you submit. Some combinations require a subscription; the Credit cost is shown in the creator.
Keep a specific image in control
When a product photo, portrait or illustration should define the starting appearance, switch to Image to Video. For motion wording you can reuse in either workflow, visit our prompt guide.
Review your result
Watch for scene consistency, subject appearance and unintended text or objects. Refine the prompt if the first clip drifts. Completed results are available in My Creations; Pricing explains the current plans.