What is Warudo?
Warudo is a Windows VTuber app built around a visual node graph. You load a VRM or Live2D avatar, wire tracking, props, and scene rules in nodes, then send video to OBS Studio or a virtual camera. The broadcast layer is the same one VSeeFace uses. Warudo adds room for a reactive show around the avatar.
The name shows up once a single tracked window is no longer the show. Warudo is deliberately larger than a free face tracker: motion and the small stage around the character, paid for with a real learning curve.
A show runtime, not a model studio
Warudo sits in the performance layer of a VTuber stack, one step past a plain tracker. You bring a finished .vrm or Live2D model. The app tracks you, renders the avatar, and can drive props and scene states. It does not sculpt meshes, paint a character from scratch, or replace VRoid Studio and Live2D Cubism. Art still happens upstream. The graph only performs what you already own.
That split is why “I installed Warudo and I still have no avatar” is the same stuck point as with any tracker. The installer is the runtime and the scene logic, not the character.
VRM and Live2D: the files Warudo expects
Warudo loads VRM 3D avatars and Live2D models in the same Windows app. A .vrm from VRoid Studio, a commission, or a licensed premade is the 3D path. A Cubism project is the 2D path. It does not play raw PNG sprite sheets the way Veadotube Mini does, and it does not author those files for you.
Match the app to the file you already have. A PNG pack will not become a Warudo puppet, and an empty node graph will not invent a model that was never exported.
What you get on stream
On a normal setup Warudo covers face tracking for VRM and Live2D: mouth, gaze, blinks, and head motion. You then hang props, hotkeys, and state changes on the node graph. Chat redeems and keyboard triggers can throw items, swap expressions, or step a gag across several beats. How far that goes depends on the model and on whether you stay on the free core or use Pro nodes.
You can build a small stage—camera moves, throwable props, multi-step bits—that a single VSeeFace window does not attempt. That control costs time. An empty graph on debut night usually looks worse than a clean tracked window. Start with one prop you will actually trigger.
Who Warudo is for
Warudo fits Windows VTubers who already stream and want reactive 3D, or mixed 2D and 3D, without opening Unity or Unreal. It is a strong second app for variety, tech, and gag formats once a talking-head tracker feels small.
It is a weak fit for a first stream this week with no model, for Mac or Linux from the developer, or for anyone who wants the smallest possible setup. Those debuts stay on VSeeFace for VRM, VTube Studio for Live2D, or Veadotube Mini for PNG until the show actually needs nodes.
Where it sits next to OBS
Think in two windows. Warudo moves the avatar and the scene logic. OBS mixes avatar, game, alerts, and audio before the stream key. Warudo does not replace OBS, and OBS does not replace tracking or the node graph. Virtual camera output is the fallback when the Warudo window flickers, or when you need the avatar in a call. OBS still owns the broadcast layout.
