This is an AI video generation comparison for
keyframes-video 
prompt:
A symmetrical, two-story blue house with white trim and a bright red door stands perfectly centered in the image The background features a bright blue sky with scattered cumulus clouds, adding depth and contrast. The house is the dominant central subject, with the tulip-lined path creating strong leading lines toward it. The camera dolly in slowly as if this is an ordinary suburban home as suddenly ((a massive human hand enters the frame, click-opening the house)) which turns out to be a dollhou...
Log in to see full prompt.
Tested: September 4, 2025
Lite is 'almost there'. It can't quite figure out how the dollhouse opens but tries to make the next best thing by just touching the house and pusing into an open door.
Tested: September 4, 2025
Pixverse was doing great with similar keyframes initially but once I've changed the end frame to a more realistic one the hand stopped appearing, not sure if coincidence. Prompt was modifid slightly to fix that but it didn't help.
Tested: September 4, 2025
Had to change original prompt because sid epanels were folding the wrong way. Opening now happens easier but man is freaking out a bit too much ))
Does the camera dolly in smoothly, as if approaching a normal house?
Does a giant, realistic human hand enter and interact with the house?
Does the house open like a dollhouse—panels hinged, swinging outward?
Are the cut-out windows on the panels aligned with the house windows?
Is there a tiny, realistic man in the top-left bedroom of the interior?
Does the tiny man react convincingly - jolting upright, gasping, yelling?
Check out the results from Kling AI (Kling 2.1) vs Dreamina AI (Seedance 1.0 Lite) vs PixVerse (PixVerse V5) vs Hailuo AI x Minimax (Hailuo 02) for similar or identical prompts side-by-side.
Pop up book with miniature diorama
