AI Floating in Space Video from One Photo
Create an AI floating in space portrait from one photo. Prepare a clear subject, describe gentle drifting, review the credit quote and track your saved video task.
Create your video
One person drifts near a curved Earth horizon. Use a small body turn, softly moving hair and a steady medium shot; keep complex hand actions out of the first attempt.
Checking your sign-in…
One JPG, PNG or WebP photo, up to 10 MB. Sign in before uploading.
Seedance 2.5 · 720p · Photo appearance reference; no motion-video input is charged.
This template requests silent output. Add permitted sound separately after downloading; no existing song or voice is requested.
Read the English motion prompt
Use @Image1 as the sole appearance reference for one person, preserving their face, hairstyle, clothing and body proportions. This is not a fixed first or last frame or a second identity. Create an original fictional zero-gravity scene with the curved Earth and distant stars behind the person. Keep a steady medium or upper-body composition with space around the head and shoulders. Show one gentle slow drift with restrained hair and clothing movement, then settle calmly. Keep the same single person recognizable and limbs connected. Optional direction may describe one small motion but must not override identity. Do not copy a specific choreography, add a second person, duplicate limbs, make hard cuts, spin rapidly, fall suddenly, hide the face with hands or promise physical accuracy. No new text, logos, dialogue, voice, music or audio.
Extra direction and the sound preference are added at submission. Model guidance does not guarantee frame-by-frame matching.
Estimated credits:47 credits
Sign in to createTurn a single-person photo into a fictional zero-gravity scene near Earth and a star field. The photo supplies the person’s appearance; your direction adjusts a small floating action. This does not follow a motion-reference video or guarantee an exact face match.
Prepare the inputs
Choose one well-lit photo of one person. Keep the face, hair and clothing visible, with space around the head and shoulders. An upper-body or full-body view is useful; a very tight face crop gives the model less room for movement. The image is an appearance reference, not a locked opening or closing frame.
How to make the video
Upload your photo, select the model and describe one gentle action, such as a small shoulder turn with drifting hair. Check the duration, frame and server credit quote, then submit once. Keep the task ID and use My creations to follow it. Refreshing only queries that saved task.
Review the result
Watch the entire clip. Check facial details, attached limbs, clothing edges and the separation between the person and Earth. Floating should be slow and readable, without a sudden fall or an unrequested camera cut.
If the face changes, use a sharper photo and reduce the action. If hands distort, keep the initial pose simple and avoid reaching across the lens. Describe one subject only; a crowded photo can introduce extra people.
Common questions
What does my photo control?
It supplies the person’s face, hairstyle, clothing and body appearance. The scene is generated around that reference. It is not a first/last frame and does not bind a second character.
Can I copy a specific space selfie or exact gesture?
This photo-based template creates an original drifting scene. It does not use a fixed motion clip, copy choreography or promise the same framing, expression or gesture as another video.
What happens after I press Generate?
The server checks the quote and saves a task before tracking the provider queue and generation. Confirmed failed tasks use the existing refund flow. If the response is uncertain, check the same task instead of submitting again.
Before saving, check the face, hair, hands and floating movement. Results vary with the photo and model.
Read this page as Markdown