Upload your idle and talking poses — or an animated bounce/blink loop — and get back a clean transparent PNG or GIF pair. No visible seam between poses, no manual masking, ready for PNGtuber Plus, Kluster, or VTube Studio.
Idle + talk PNGs, or a bounce/blink loop as an animated GIF
Flood-fill follows anime line art exactly, matched across every pose
Transparent PNG or GIF, ready for PNGtuber Plus, Kluster, or OBS
Common PNGtuber Setups
Upload idle and talk poses together for a matched, seamless cutout
Single-image background removers process each pose independently, so the mask boundary drifts between idle and talk. Batch processing keeps them matched.
See also: VTuber & overlay assets and the guide to why PNGtuber loops keep a faint background.
A PNGtuber is a streamer avatar made from one or two static images (usually a mouth-closed and mouth-open pose) that swap based on your mic input, instead of a fully rigged, animated Live2D or 3D VTuber model. It's a much faster and cheaper way to have an on-screen character — the only real requirement is that both poses have a clean transparent background.
Most PNGtuber apps (like PNGtuber Plus, Kluster, or VTube Studio's PNGtuber mode) want at least two images: idle/mouth-closed and talking/mouth-open. Upload both as a batch and they'll come back with identical, consistent transparent backgrounds.
Yes — upload an animated GIF for a breathing/bounce idle loop or a blink cycle, and every frame gets the same transparent-background treatment so the loop plays back cleanly with no flicker.
That's usually leftover fringe from a single-pass background remover, or a mismatch between the mask on your idle pose vs your talk pose. Batch-processing both poses together in one job keeps the cutout boundary consistent across both, which is what removes the visible seam.