Make an AI video clip from an idea
Goal: get a short clip of a scene you describe in words.
Recipe: AI video clip from your idea
Inputs. Your idea, an optional Mood and look, How much effort and Shape. Effort sets length and takes together: Draft is 6 seconds, one take; Standard is 10 seconds, one take; Best of two is 10 seconds, two takes. Under More options you can pick the video model. You need a video-generation provider in your workspace. No person from your project is shown, so this is not the way to make an avatar; for that see Avatar explainer.
What it costs. It uses credits. Draft costs least and Best of two most. The step has a spend cap, and the price shows on the run page before you start.
What you get. One clip, or two to choose from, saved in the project.
If it fails.
- The run stops asking for a provider: add a video-generation provider to your workspace.
- Faces or readable text look wrong: describe a scene without people or words.
- The look is off: change Mood and look and try Draft first, then a higher effort.
Permission to send media. This recipe sends only your description. If you later add a step that sends your own footage, that step stays off until you switch it on in the app, or ask the assistant to do it and approve the card that names the step.