Wan AI 2.2 Animate Replace: Video Character Replacement Online
Bottom line
Swaps the performer in existing footage with any character from one reference image. Bring a source video and a character image to the Animate Replace tool — VdoBloom prices the swap in credits up front and refunds automatically if a run fails.
Most video models generate new footage; Wan AI 2.2 Animate Replace edits who is in footage you already have. The workflow takes two inputs — a source video and a character image — and re-renders the clip with your character performing the original motion. Camera moves, timing, and scene layout carry over from the source; the person on screen changes.
That makes it less a generator and more a recasting tool. The Wan family comes from Alibaba, whose open releases of earlier versions made it one of the most widely adopted video model lines, and 2.2 is the generation this replacement variant is built on. Output resolutions are 480p, 580p, and 720p — working quality for social clips, previz, and drafts rather than broadcast masters.
Animate Replace carries a HOT content label, VdoBloom's most permissive tier, so it tolerates material that stricter models refuse. It is rated 4.3 by users. Each job is billed in VdoBloom credits, with the price shown before the render begins, so recasting experiments are costed before you commit a source clip.
What creators use Wan AI 2.2 Animate Replace for
- Recast a filmed performance with a brand mascot using a single character image.
- Swap yourself into stock or archive footage for a channel intro without a reshoot.
- Preview how different character designs read in motion using one source clip.
How to use Wan AI 2.2 Animate Replace on VdoBloom
- 1Open the Animate Replace tool and select Wan AI 2.2 Animate Replace.
- 2Upload the source video and the character image you want inserted.
- 3Pick the output resolution, review the credit cost, and generate.
Frequently asked questions
What inputs does Wan 2.2 Animate Replace require?
Two things: a source video and a character image. The model re-renders the video with the pictured character performing the original motion — it is a video-to-video replacement workflow, not a text-prompt generator.
Does the output keep the original camera work and timing?
Yes — that is the point of the replacement approach. The source clip's motion, framing, and timing carry over; what changes is who appears on screen, not what happens in the shot.
What output resolutions are available?
480p, 580p, and 720p. That positions it for social content, previsualization, and drafts rather than broadcast delivery, so plan your pipeline accordingly.
How permissive is Animate Replace's content filter?
It carries the HOT label, VdoBloom's most flexible tier, meaning filtering is considerably more permissive than CENSORED-tier models. Standard platform rules still apply, but creative directions refused elsewhere generally run here.