The output ignores the prompt
Shorten the prompt and put the subject, action, and camera move at the beginning.
TEXT TO VIDEO AI
Quick answer
Text to video generates a short video from a written prompt. Start with a focused scene idea, then choose model, aspect ratio, duration, resolution, sound, and credits before generating.
Input requirements
Match the source type to a conservative first motion target. That improves the keeper rate before spending credits on longer clips.
| Input | Best for | Avoid |
|---|---|---|
| Product or ad prompt | Clear product, camera move, lighting, and scene style | Too many products or contradictory camera directions |
| Social video prompt | Vertical framing, strong opening action, simple subject | Long scripts or many scene changes |
| Cinematic prompt | One location, one subject, clear mood, and camera language | Multiple cuts or complex story beats in one generation |
Failure reasons
Shorten the prompt and put the subject, action, and camera move at the beginning.
Add concrete visual details such as location, light, lens feel, and subject behavior.
Reduce the prompt to one scene and one motion goal before adding style.
FAQ
No. Text to video starts from a prompt. Use image to video or reference to video when you need a source visual.
A good prompt names the subject, action, environment, camera movement, lighting, and style without overloading the model.
Yes. The workbench includes product ad, social hook, and cinematic scene templates to help you start faster.
Related pages
Use one still image as the visual source and prompt only the motion.
Open image to videoREFERENCEGuide the subject or style with a reference image plus prompt.
Open reference to videoPHOTOAnimate one real photo with a more conservative photo-first workflow.
Open photo to video