How to Make a Rumpelstiltskin AI Video With Your Photos

The whole point of AI Rumpelstiltskin is that you do not need editing software, a prompt-engineering degree, or more than two minutes. Here is the entire workflow.

Step 1: Pick your photos

You need one clear photo of the person (or pet) who will play the little man. Front-facing, good light, face unobstructed. A second photo for the princess character is optional but recommended — the two-character version is the format at its best.

Step 2: Upload and describe

Drop the photos into the generator. Optionally, describe the scene: outfits, location, a finishing move. "Gold sequin tuxedo in a neon nightclub, ends with a shush" is a classic. You can also leave it empty and let the barn do the talking.

Step 3: Choose resolution and length

Videos render in 16:9 landscape at 720p, 1080p, or 4K, and run 3 to 9 seconds. Short versions loop better in group chats; the full 9 seconds carries the complete dance.

Step 4: Generate and post

Press generate. Credits are deducted only when the render starts — a 9-second video costs 60 credits, a 3-second one 20 — and if the render fails they come back automatically. Your video lands in about two minutes. Add the trending sound in your app, label it AI-generated, and post.

Tips

• Head-to-waist framing preserves the outfit.
• Skip mirror selfies — the phone shows up in the render.
• One subject per photo; use the second slot instead of combining faces.

That's it. Two photos in, one barn video out.