Two modes, one framework. Animation Mode drives a character image with the motion and expressions from a reference video. Replacement Mode inserts your character into existing footage — with lighting and color tone matched automatically.
Animate a character image with our online generator today — full reference-video mode workflows are coming to the platform soon.
Open the GeneratorAny character image — mascot, illustration, product figure — becomes the performer.
Animation Mode borrows body movement and facial expression from a reference clip; no mocap suit required.
Replacement Mode drops the animated character into real footage, auto-matching lighting and color tone via its Relighting LoRA.
Facial features are transferred implicitly — smiles, glances and speech rhythms read as genuine performance.
Spatially-aligned skeleton signals keep limbs and posture faithful to the source performance.
Replace any speaker in interviews, reels and dubbed clips while preserving mouth shapes matched to original audio.
Produce unlimited mascot content where the character stays perfectly on-model across every scene.
Wan Animate is Alibaba's unified character animation and replacement framework built on Wan image-to-video models. It animates a character image using a reference video's motion (Animation Mode) or replaces a character within existing footage (Replacement Mode).
Animation Mode creates new footage of your character following a reference performance. Replacement Mode takes existing footage and swaps the on-screen character for yours, matching lighting, color tone and mouth shapes to the original.
No. A simple phone recording of someone performing serves as the motion reference — Wan Animate extracts both body movement and facial expression from it.
Yes — those are two of the most popular uses: driving a consistent avatar for streams, and re-voicing clips while keeping believable lip synchronization.
