Remove the background from a screenshot — on-device, on Windows.
One click isolates your subject with on-device AI. No upload, no cloud, no account — it works offline after a one-time model download, so your screenshots never leave your machine.
On-device, not in the cloud
ShotsGlow runs a local AI model (BiRefNet) to separate your subject from its background. The whole path — decode, inference, composite — happens on your computer; no network call is made and no pixels are sent anywhere. It uses your GPU through DirectML when it can, and falls back to the CPU automatically.
That matters for screenshots specifically: they often contain a customer dashboard, an unreleased UI or a client's data. Removing a background shouldn't mean uploading that to someone's server.
How it works
- Click Remove background — the subject is isolated automatically
- Drop it onto a gradient backdrop, a solid, or keep it transparent
- One-click Restore original if you change your mind
- Pair it with the markup kit, redaction and 30 export sizes — same app
Honest scope
On-device background removal exists in a few point tools. What's different here is that it's native to Windows and built into the app you already captured, annotated and polished in — not a separate upload-and-download round trip. It's one feature of ShotsGlow, not the whole product.
Does it upload my screenshot to remove the background?
No. The AI model runs on your machine — image pixels never leave your computer. There is no cloud service and no account.
Does it work offline?
Yes. After a one-time model download (~928 MB), background removal runs fully offline on your device.
Does it need a special GPU?
It uses DirectML GPU acceleration when available and automatically falls back to the CPU, so it works on a wide range of Windows PCs.
How is it different from remove.bg or Canva?
Those are cloud services — you upload your image to their servers. ShotsGlow does the cutout locally, inside the same app you captured and polished in.