Generate your first video

Create your first video clip on the board.

Before you start

Steps

  1. (Optional) Select an image on the board to use as a starting frame.
  2. Describe the scene in the agent input. For example: "A slow dolly shot through a sunlit forest, leaves swaying gently."
  3. (Optional) Type @ to pick a specific video model — or let the agent choose for you.
  4. Press Enter. Video generations take 30–120 seconds.

Tips

  • Without a selected image, you get a text-to-video generation. With a selected image, the model uses it as a starting frame.
  • Short clips (4–5 seconds) generate faster and are great for testing.
  • You can queue up to 5 generations at once.
  • To learn more about picking models, see How to pick a model.

Next step

Create your first composition — or explore the full Guides section.

On this page