Replace (character replacement) focuses on environment integration. It keeps the original video background, camera movement and lighting, and only replaces the person with your uploaded character, which is ideal for visual effects and compositing.
Move (motion-driven) focuses on motion replication. It mainly cares about making your static image move; the background may be simplified or regenerated, with emphasis on accurate body motion and facial expressions, which is ideal for VTubers or animated stickers.

Image: we recommend a high-resolution, front-facing, unobstructed full-body character image with clearly visible limbs.
Video: choose clips where the subject stands out clearly from the background and lighting is even. Avoid overly crowded scenes or overlapping characters.

Yes. You can record a real-person talking or performance video, then upload your virtual avatar illustration and use Move mode to quickly generate VTuber-style content, without buying expensive motion capture equipment.

In most cases it will not. WeryAI uses advanced temporal consistency algorithms to greatly reduce flicker. In extreme motions or very complex backgrounds, minor artifacts may appear, so we recommend testing with shorter clips first.