The short answer

What is a VTuber?

A VTuber is a streamer or video maker who appears on screen as a drawn character instead of on camera. The character is driven live by the person behind it — usually by a webcam reading their face, sometimes just by their microphone — so when they talk, it talks, and when they look surprised, so does it.

VTuber, in one line

Short for virtual YouTuber. A performer whose on-screen body is an animated character they control in real time.

  • It is nota filter over your face, or a pre-rendered cartoon.
  • It is notanonymous by definition — plenty of VTubers are publicly known.
  • It is notone piece of software. It is a chain of four, described below.

The longer answer

The word is a contraction of virtual YouTuber, and it has outgrown both halves. VTubers stream on Twitch as much as they post to YouTube, and almost none of them are virtual in the sense of being fictional — there is a person at a desk, talking, and the character on screen is doing what their face is doing a few milliseconds earlier.

The useful way to think about it is that the avatar is a costume rather than a character in a story. A VTuber playing a horror game jumps when the game does. The interest is not that the character is animated; it is that it is animated by someone, right now, and the audience can tell the difference immediately.

The format started in Japan in the late 2010s and the largest agencies are still Japanese, but the tools stopped being specialist some years ago. The practical question for anyone reading this is no longer whether it is possible — it is which of the four pieces below you buy, which you make, and which you skip.

What a VTuber setup is made of

Every setup has these four, from a hobbyist's first stream to an agency debut. What separates them is how much each piece cost.

Only the second piece really varies. A rig and an expression sheet are two different answers to the same problem, and which one a tool uses decides almost everything else about it.

PNGTuber vs VTuber

The two words get used as if one were a cheaper version of the other. They are closer to two different trades, and the cheaper-looking one is the right answer for a lot of people.

PNGTuberVTuber
What movesA whole image swaps for another one. Usually two: mouth shut, mouth open.Parts of one image move independently — jaw, eyes, head, sometimes the body.
What drives itYour microphone. Cross a volume threshold and the image changes.Your webcam, reading your actual face.
What it costs to set upTwo drawings and a free tool. Some people are live in an afternoon.A rigged model, which is a commission or several weeks of learning.
What it asks of youNothing. No camera, no calibration, no lighting to think about.A webcam pointed at your face, lit well enough to read, for the whole stream.
Who picks itPeople who want to be on stream this week, and people who would rather not be on camera at all.People who want the avatar to carry expression the way a face does.

A PNGTuber is a streamer represented by a still image that swaps between two states — quiet and talking — when their microphone crosses a volume threshold. There is no camera and nothing to rig. People choose it because it works, not because they could not afford the alternative.

A VTuber is the same idea with a face behind it. The avatar carries expression because your expression is driving it, which is worth a great deal on a talking stream and almost nothing if you stream with the lights off.

The honest test is whether you want to be on camera. Not on screen — the avatar is on screen either way — but on camera, lit, for the length of a stream. If the answer is no, PNGTubing is not the compromise. It is the correct choice, and it will still be the correct choice in a year.

The other words you will hit

So how do you actually become one?

The equipment, the choice between the routes above, and what the first stream needs — laid out with what each part costs in time and money.

Read the guide

GetGuise is one way to be either.

A drawn character with a sheet of expressions, driven by your webcam in a window OBS and Streamlabs capture. Free to download, with the characters already in it.

See what it does