Image to VRM — a 3D avatar from a picture of your character

For characters a selfie can't capture: anime, art, a game character. We redraw the character in a T-pose, build it in 3D and rig it for VTubing. You get a VRM and a GLB. The face does not move — no blinking, lip sync or expressions; the head and body follow tracking.

Try it with your pictureFree T-pose preview · $4.99 per avatar · 15 to 25 minutes

Examples

Three pictures and what came out of each. Nothing was fixed by hand: the T-pose is drawn from the picture, the 3D model is built from the T-pose, and the last two frames are that model, rigged, with its arms lowered.

The picture: an anime drawing of a girl with a pink bob haircut in a yellow hoodie
1. The picture
The same character redrawn full-body in a T-pose: a girl with a pink bob haircut in a yellow hoodie

2. Free T-pose preview

The rigged 3D avatar built from the picture, front view: a girl with a pink bob haircut in a yellow hoodie

3. The 3D avatar

The same 3D avatar turned three-quarters: a girl with a pink bob haircut in a yellow hoodie

150,000 triangles · 16.4 MB VRM

The picture: an anime drawing of a young man with spiky dark blue hair in a black bomber jacket
1. The picture
The same character redrawn full-body in a T-pose: a young man with spiky dark blue hair in a black bomber jacket

2. Free T-pose preview

The rigged 3D avatar built from the picture, front view: a young man with spiky dark blue hair in a black bomber jacket

3. The 3D avatar

The same 3D avatar turned three-quarters: a young man with spiky dark blue hair in a black bomber jacket

150,000 triangles · 16.4 MB VRM

The picture: an anime drawing of a young woman with a mint green side ponytail in a navy blazer
1. The picture
The same character redrawn full-body in a T-pose: a young woman with a mint green side ponytail in a navy blazer

2. Free T-pose preview

The rigged 3D avatar built from the picture, front view: a young woman with a mint green side ponytail in a navy blazer

3. The 3D avatar

The same 3D avatar turned three-quarters: a young woman with a mint green side ponytail in a navy blazer

150,000 triangles · 14.5 MB VRM

How it works

  1. 1

    Free preview

    We redraw your character full-body in a T-pose. About 15 seconds, 3 previews a day.

  2. 2

    You approve it and pay $4.99

    One payment, one generation. Nothing is charged before you press the button.

  3. 3

    We build and rig the 3D model

    15 to 25 minutes. You can close the page: we email you when it is ready.

  4. 4

    Download the VRM and GLB

    The download is included in the $4.99.

Before you pay

  • The face does not move: no blinking, lip sync or expressions. The head and body follow tracking.
  • There is no refund once the avatar is delivered.
  • If we can't deliver a file, the $4.99 is refunded automatically.
  • You confirm you have the right to use this picture.

Common questions

What kind of picture works?

A drawing or render of one character: anime, illustration, a game character, a mascot. PNG, JPEG or WEBP up to 15 MB. A portrait is enough — the rest of the body is drawn for you in the free T-pose preview, and you decide from that preview whether to go on. A full-body picture leaves less to guess.

Does the face move?

No. The avatar has no blinking, lip sync or expressions: the eyes and mouth are part of the texture. The head and body follow tracking. This is said before payment, and you tick a box to confirm you have read it.

What do I get?

A VRM and a GLB of the same model: one mesh of at most 150,000 triangles with its textures, a humanoid skeleton with all 15 bones VRM requires, standing in a T-pose. The VRM is version 0.x and loads in VSeeFace, Warudo, VNyan and 3tene.

How much is it and can I get a refund?

$4.99 per avatar, paid once, after you have seen the free preview. The preview is free, three a day. If we cannot deliver a file, the $4.99 is refunded automatically. There is no refund once the avatar is delivered.

How long does it take?

The T-pose preview takes about 15 seconds. After payment the 3D model, the skeleton and the VRM take 15 to 25 minutes in total. You can close the page: we email you when the avatar is ready.

I want an avatar that looks like me, with a face that moves.

That is the selfie creator: a photo of your face becomes a photorealistic avatar with blinking, lip sync and expressions, and you restyle hair and outfit in the editor.

Make an avatar from a selfie →

My character already exists as a 3D model.

Then nothing needs generating. The converter maps the skeleton of an FBX, GLB, Blender or MMD model to the VRM humanoid and converts its materials.

Convert a model you already have →

See your character in a T-pose first

The preview is free and takes about 15 seconds. You pay only if you like what it shows.

Try it with your picture