Skip to main content
Requires Premium.
Image-to-Video takes an existing image and animates it. Maximum duration is five seconds.

How to use it

  1. Generate an image using the Realism model in the image generator
  2. Save the image to your gallery
  3. Open the video generator and start or open a Session
  4. Select Image-to-Video
  5. Select your image from the gallery
  6. Write a motion prompt describing what should move and how
  7. Select a resolution matching your source image
  8. Click Generate

Resolution

The video resolution must match the source image resolution. Select the same aspect ratio in the video generator that you used when generating the image.
Source image ratioUse in video
1:11:1
7:97:9
9:79:7
5:125:12
12:512:5

What makes a good first frame

Not every image animates cleanly. These factors produce better results:
  • Mid-shot or full-body framing — extreme close-ups leave little room for natural motion
  • Simple background — busy backgrounds produce visual noise when the model separates foreground from background motion
  • Neutral or dynamic pose — fully static poses produce subtle motion only; a slightly dynamic pose gives the model more to work with
  • Soft or flat lighting — hard shadows shift unnaturally across frames

Token cost

Image-to-Video uses tokens. See /premium for current rates.

Common issues

The first frame may have a complex background or extreme pose. Regenerate the first frame with a simpler background and a more neutral pose, then try again.
Only images generated with Realism support image-to-video. Regenerate the first frame using Realism.
Add more explicit motion terms to the prompt — walking forward, turning head, hair blowing, camera slowly zooming in. See Video Prompting.

Workflow

Full step-by-step from first frame to finished video.

Video Prompting

Motion and camera terms that produce consistent results.

Enhanced Video

Upscale and sharpen the output after generation.

Sessions

Keep your video generations organized.