Tutorial

Images in the terminal

Sinclair renders sixel graphics: programs print an escape sequence, and an actual image appears in the grid — composited on the GPU, anchored to the text it arrived with, scrolling away with the rest of your output. File previews, plots, thumbnails, QR codes: if a tool can speak sixel, it draws right in your terminal. This walkthrough renders your first image, explains how tools detect support automatically, and wires up a couple of everyday uses.

1. Render your first image

Grab either of the two standard converters and point it at any image:

# with libsixel
brew install libsixel        # apt/dnf: libsixel-bin
img2sixel photo.png

# or with chafa
brew install chafa
chafa -f sixels photo.png

The image prints like output — because it is output. Run it again, scroll back, and both copies are exactly where they happened in the history. Since sixel is just bytes on the wire, this works over ssh too: run img2sixel on a remote host and the picture renders in your local Sinclair.

2. How tools know they can do this

You don't have to configure anything. Sinclair advertises sixel in its DA1 (Primary Device Attributes) response and answers XTSMGRAPHICS and XTWINOPS queries, which report the pixel dimensions of the cell grid. Programs that probe the terminal — the way notcurses, timg, or matplotlib's terminal backends do — light up sixel support on their own and size images to your actual cells.

3. Previews in your file manager

Terminal file managers make the best everyday use of this. yazi shows sixel previews out of the box when the terminal advertises support. For lf or ranger, set the previewer to chafa -f sixels and browsing a directory of screenshots becomes an image gallery — without leaving the keyboard.

4. Plot from Python

For quick data exploration, a sixel matplotlib backend (e.g. matplotlib-backend-sixel) prints figures straight into the session:

pip install matplotlib-backend-sixel

python -c "
import matplotlib; matplotlib.use('module://matplotlib_backend_sixel')
import matplotlib.pyplot as plt
plt.plot([1, 4, 2, 8, 5]); plt.title('right here in the terminal'); plt.show()
"

The figure appears inline, scrolls with your session, and is captured in the scrollback next to the code that produced it — a lab notebook for free.

Where to next