sss_cli

sss_cli (binary: sss) is the flagship screenshot tool. Pick a region, a window, or a monitor; annotate it; save or copy. Wayland and X11 first-class; macOS via the native ScreenCaptureKit path.

sss_cli annotation overlay — arrow, shapes and blur on a captured region
Annotation overlay — arrows, shapes, blur, text

Multi-monitor, no seams

--area drags across your whole virtual desktop — pull a rectangle from one monitor to another and the captured PNG stitches the frames together at the right scale, no visible cut. Works on Wayland (via wlr-screencopy per output + virtual-desktop math), X11 (RANDR 1.5 GetMonitors) and macOS (CGGetActiveDisplayList).

phone photo of two physical monitors with a single sss selection rectangle spanning the gap between them
Real hardware — one selection across two monitors

At a glance

  • Three target modes: screen, window, area — each with an interactive picker or a direct selector.
  • Annotation overlay: shapes, arrows, text, pen, blur, pipette.
  • Optional OCR layer (oar-ocr) to copy text directly from images.
  • Compositing options: cursor on/off, decoration on/off, border on/off.
  • Backends auto-detected: wayland, portal, x11, windows, macos.

When to use it

GoalFlag combination
Region select with editorsss --area
Window pickersss --window
Whole screen, no UIsss --screen --current
Force portal backendsss --capture-backend portal
Disable OCRsss --ocr=false

OCR

The annotation overlay can pipe the captured (or selected) region through an OCR engine and hand the recognized text straight back to the clipboard. Powered by sss_ocr with hardware-aware defaults (CPU / CUDA / TensorRT / CoreML / DirectML / OpenVINO / WebGPU).

sss_cli OCR — text recognition boxes over a captured region
OCR — select, recognize, copy

Disable with --ocr=false or pin a backend with --ocr-ep cuda|tensorrt|coreml|directml|openvino|webgpu.

Walk through the getting started guide next, or jump to flags for the full reference.

pages

In this section