Here is a moment every researcher knows. You're reading an article and hit the figure — the chart that makes the whole argument, the table with the numbers you'll want to quote, the diagram that explains in one image what the text takes ten paragraphs to say.
Text is easy to move into Gemini Notebook. Images are not. The usual routine: full-page screenshot, find it in your Downloads folder, crop it in an editor, drag it into Gemini Notebook, delete the leftover file. Five steps, two apps, and by the end you've lost your reading flow entirely.
Web Snip collapses all of that into one gesture.
How to snip anything
1Right-click anywhere on the page and choose "Web Snip this page (Notebook OS)" from the menu.
2The page dims and your cursor becomes a crosshair. Drag a rectangle over the chart, table, quote, or diagram you want. (Press Esc to cancel.)
3The capture appears instantly in the Web Snip card of your side panel — cropped exactly the way you dragged it, labeled with the page title and URL so you always know where it came from.
From there you have three buttons: Copy puts the image on your clipboard, Save snip to Gemini Notebook uploads it into the notebook you pick, and New starts another capture.
Why the right-click? A privacy feature, not a workaround
Most capture extensions demand permission to "read and change all your data on all websites" the moment you install them — standing access to every page you will ever open, whether you capture it or not.
Web Snip is built on the opposite principle. Notebook OS ships with access to Gemini Notebook and its research sources only. When you right-click a page and choose Web Snip, Chrome grants access to that one page, for that one action — a mechanism Chrome designed exactly for this. No standing permissions, no background access, nothing to revoke later because nothing persists.
What snips are good for in Gemini Notebook
Gemini Notebook's AI reads images. That's what makes Web Snip more than a screenshot tool — every snip becomes a source the AI can reason about:
- Chart from an article → snip it, then ask Gemini Notebook to explain the trend or pull the numbers into a table.
- Table in a paywalled report → snip the visible table and query it alongside your other sources.
- Architecture diagram from documentation → snip it and have your notebook explain each component.
- A striking pull quote → snip it with its styling intact, so the artifact keeps the original's voice.
Chain it with the rest of Notebook OS: snip three charts while reading, add your own take with Capture Spoken Thoughts, then run a Studio Recipe to turn all of it into a report or deck.
The honest fine print: PDFs
One thing Web Snip cannot do: capture Chrome's built-in PDF viewer. That's a Chrome platform restriction — extensions aren't allowed to reach into it, and we'd rather tell you plainly than pretend otherwise. If you try, Notebook OS says so directly and suggests the fastest alternative: your system's own screenshot tool (Win + Shift + S on Windows, ⌘ + Shift + 4 on Mac), then add the image to your notebook in Gemini Notebook.
Losing a snip is hard by design
Every capture is kept locally in your browser until you've saved it — close the panel, switch tabs, come back tomorrow: your last snips are still there in the Web Snip card, waiting. And like everything in Notebook OS, they never touch an external server. The image goes from your screen to your browser to your Gemini Notebook — and nowhere else.
The best research material is rarely plain text. Now the charts and diagrams can come along too.