A VTUBER MODELING EXPERIMENT

Can Astra build
a VTuber model?

This is our answer so far: a silver-fox character with a partially working facial rig in VTube Studio. The face is taking shape. The body is still a work in progress.

THE MODELActual VTube Studio capture

FACE STUDIES

Small changes. Different emotions.

Six static combinations with the existing hair and ears.

Model captures are real VTube Studio output. Face artwork and mouth studies are separate static previews of newer parts, awaiting native rig integration.

Explore supporting artwork & production files

LOOK UNDER THE LAYERS

Behind the model

Eyes, brows, shading, and mouth studies on a shared canvas.

BUILT WITH ASTRA IN CODEX

The experiment is
the character itself.

We used Astra as the production agent to work through artwork, separated parts, facial rigging, and repeated fixes. It used image generation, scripts, Cubism, and VTube Studio; Luna subagents helped with focused tasks.

Create & organize

A silver-haired fox character, developed from reference art into layered artwork and an actual native model.

Compare & refine

Eye and mouth movement and expression work are partially implemented. New facial artwork is still being integrated and refined.

Rig & test

The body, arms, joint shapes, and movement naturalness still need work. A loaded model and individual expression checks are progress, not a finished full-body rig.

This entry documents a VTuber model made with Astra. It does not connect an AI chatbot to the character or offer one-click modeling. The images shown here are labeled as native model captures or source-art studies; full rigging and final camera acceptance are still in progress.

Can Astra builda VTuber model?