Wireframe first.
Then the scene.
Write the scene in a sentence. ImageAnimator sketches the structure, fills it in, and hands back a short vertical clip.
"moon over a ridge, mist in the valley, slow pan"
How it works
Three steps
Subject, place, camera move. One sentence is the whole input.
The structure is laid out first, then rendered and set in motion at 9:16.
Saved with its sentence. Change a word, animate it again.
Output
What you get
Framed while rendering, never cropped.
Longer stories work as several clips.
Saved with every take.
Runs in the browser you have.
FAQ
Before you ask
Will it match my sentence exactly?
No — the line steers the structure rather than specifying every frame. Two runs differ; concrete nouns get you closer.
How long can a clip be?
Most land between five and fifteen seconds. For longer, render several and place them one after another.
Can I use it commercially?
Yes, within our terms. Real people's likenesses and copyrighted characters are declined at generation time.
What does it cost?
Metered by clip length, shown before a render starts; nothing runs without you confirming it.