Image to video AI: turn any photo or drawing into a video
Image to video starts from a picture you already have — a product photo, a character illustration, a landscape, a meme — and animates it. You upload the image, describe what should move and how the camera should behave, and the video model turns it into a clip that begins exactly on your image.
Because the clip starts on your frame, image to video gives you much more control over composition and character design than text alone.
How it works
- 1
Upload an image
JPG, PNG or WebP up to 5 MB. A clean, well-lit image with one clear subject animates best.
- 2
Describe the motion
Write what moves and how: "the waves crash against the rocks, camera slowly pushes in, mist drifts across the frame".
- 3
Pick a model and format
Choose a video model that supports start frames and the aspect ratio you need. The clip length follows the model (for example up to 15 seconds on HeyGen, Kling or Wan 2.7).
- 4
Download
The clip is saved to your library as an MP4, with the model’s own sound when the model generates audio.
Writing motion prompts
- Describe one main action. "She turns and walks toward the door" animates better than five things happening at once.
- Say what the camera does: static, slow push in, pan left, orbit, handheld.
- Mention secondary motion for life: hair in the wind, steam, flickering light, drifting snow.
- Avoid asking for text or logos to appear — video models still render lettering poorly.
Good images to start from
Images with a clear subject, room around it and consistent lighting give the most natural motion. Very busy collages, tiny subjects, heavy text and extreme crops make it harder for the model to decide what should move. If your image is landscape but you want a vertical video, the clip is cropped to fill the frame, so keep the subject near the centre.
Frequently asked questions
Can I make a longer video from one image?
Image input makes a single clip up to the model’s maximum length. For longer videos, use text input: ShortsMachine generates a key image for every scene and animates each one, then stitches them together.
Will the clip keep my character’s face?
The clip starts on your exact image, so the first frame matches. As the motion progresses, models can drift slightly; shorter clips and gentle motion keep the subject closest to the original.
Why was my image rejected?
Video providers run content checks on start images. Images with graphic violence, nudity or real public figures can be refused. Try a different image or describe the scene with text instead.
Make your first video in a few minutes
Type an idea and ShortsMachine writes the script, generates every scene, adds a voiceover and captions, and stitches it into one video — up to 5 minutes long.
Start your 3-day free trial