AI photo to video
AI depth and video encoding run in this tab. Your photo is never uploaded.
Depth Anything V2 · Apache-2.0……
The camera moves through the photo's AI depth.
How to use this tool
Make a short 3D video from one photo in this tab.
Add a photo
Drop or paste a picture. AI works out its depth in a few seconds and the preview starts moving.
Choose the move
Push in, pull out, orbit, pan or rise. Set the strength, length, shape (9:16 for Reels and Shorts) and quality.
Download the MP4
The video is encoded in the browser and saved as an MP4 you can post anywhere.
FAQ
How does photo to video work?
An AI model (Depth Anything V2) estimates how far each part of the picture is. A virtual camera then moves through that depth, so near things shift more than the background, like a short 3D shot.
Is my photo uploaded?
No. The model downloads once (about 25–100 MB, depending on your device) and runs in this tab. The video is encoded in the browser too.
Can it make a video from text?
No. It animates a picture you already have: a photo, a drawing or an AI image. People and objects do not move on their own; the camera moves.
Which format and size is the video?
MP4 at 30 fps, 3, 5 or 8 seconds, 720p or 1080p (never more than the photo's own pixels), in the photo's shape or cropped to 9:16, 1:1 or 16:9. It plays on phones and uploads to TikTok, Reels and Shorts.
Why do edges look stretched?
The camera reveals a little of what was hidden behind near objects, and the tool fills it from nearby pixels. Pick Subtle, or a photo with a clear subject and a plain background.