Skip to main content
Skip to content
Generate a video from text

Generate a video from text

Turn a written prompt into a generated video clip.

Before you start

This guide creates a clip from a written description. In Storyboard, open a shot and choose Animate, then Text to video. Or open Generate videos and choose Text to video there. Video generation needs Basic or above. Seedance 2.5, Seedance 2.0, and Seedance 2.0 Fast need Plus or above. Choose Image to video if you want a specific image as the starting frame.

Steps

  1. In Storyboard, choose Animate on the shot, then Text to video. In Generate videos, choose Text to video. Both open the same kind of model and prompt settings.

  2. Select a Video model from those available for this workflow. Changing the model can change the required inputs, durations, output settings, and credit cost.

  3. Write the Prompt. Describe the subject, action, setting, and camera movement. For example: “A pizza chef tosses dough as a small dog watches; warm morning light, slow camera push-in.”

  4. Optionally open Director for reusable visual and camera direction. Check how Element tags are supported by the selected workflow; a label such as Tags · Descriptions only means tags provide text descriptions rather than reference images.

  5. Choose the available Duration, Resolution, and Aspect Ratio. Review audio options when the selected model offers them and check the current cost.

  6. In Storyboard, choose Animate shot. In Generate videos, choose Create video. Read the credit total on the button first. A Storyboard clip appears on the shot when it finishes. A Generate videos clip appears in the gallery and in My Assets → Videos.

Expected result

The completed clip becomes the shot’s visual automatically. Generated video is also saved in My Assets → Videos, and you can use it in Editor.

Duration, cost, and plans

Duration is in whole seconds. The credit charge is the output length, rounded up to a whole second, times the model’s rate for the selected resolution. The button shows that total before you submit. Rates below are credits per output second.

  • Seedance 2.5 (Plus and above): 4–30 seconds. 20 at 480p, 40 at 720p, 100 at 1080p.
  • Seedance 2.0 and Seedance 2.0 Fast (Plus and above): 4–15 seconds. Seedance 2.0 is 20/40/100 at 480p/720p/1080p, and 200 at 4K. Fast is 17 at 480p and 35 at 720p.
  • Seedance 2.0 Mini (Basic and above): 4–15 seconds. 7 at 480p and 14 at 720p.
  • Kling 4.0 (Basic and above): 3–30 seconds. 48 at 720p and 72 at 1080p. Generate audio is the intended sound control and starts on.
  • H3 Original, shown as Hailuo 3: 4–15 seconds. H3 Max and H3 Max Turbo: 1–15 seconds. Wan 3.0 and Wan 3.0 Prime: 2–30 seconds. FLUX 3 and FLUX 3 Max: 5–20 seconds. P-Video 2: 1–20 seconds. P-Video 2 Pro: 5–15 seconds. Gemini Omni Flash 1.1: 3–10 seconds. Happy Horse 1.1: 3–15 seconds. Grok Imagine Video 1.5 and Grok Imagine Video 1.5 Lite: 1–15 seconds. PixVerse V6 and PixVerse C1: 1–15 seconds. PixVerse V6 Max: 5, 8, 10, or 15 seconds. LTX 2.5 Pro: 6, 8, or 10 seconds. LTX 2.5 Fast: 2, 3, 4, 5, 6, 8, 10, 12, 14, 16, 18, or 20 seconds. Runway Gen-4.5: 5 or 10 seconds.
  • Models with a Generate audio switch charge the with-audio rate when that switch is on. The button is the place to confirm the rate for the model and resolution you selected.

Tips

Keep a short clip focused on a manageable action. A model can interpret complex instructions differently than expected. Preview motion, framing, and audio before generating more versions.

Documentation