A single element in Chrome's page representation. Each node carries a type (one of 21, from headings and links to form controls and dialogs), bounding box coordinates, text styling, accessibility metadata, and an ID that the AI uses to reference or interact with that specific element.
A single element in Chrome's structured page representation. When Chrome extracts a page for Gemini, every visible element becomes a content node in a hierarchical tree.
Each node is assigned one of 21 types: Root, Container, Text, Paragraph, Heading, Anchor, Image, SvgRoot, Canvas, Video, Form, FormControl, Table, TableRow, TableCell, OrderedList, UnorderedList, ListItem, Iframe, DialogModal, and DialogModeless. The type determines how the element is treated — a FormControl is something the AI can fill in, an Anchor is something it can click, a Heading signals document structure.
Beyond type, each node carries geometry data — three kinds of bounding box (outer, visible, and fragment) in viewport pixel coordinates — text styling (font size ratio, emphasis, color), accessibility metadata (focusable, tabbable, disabled, 16 possible clickability reasons), and landmark roles (header, nav, main, footer, etc.).
Not every node in the full rendering tree reaches Gemini. Structural containers and layout wrappers are filtered out. For a typical page, Chrome might extract 471 nodes from the rendering tree but only pass 198 tagged nodes in the structured Markdown — the ones Gemini can read or interact with.
