Web Node
Capture a webpage's screenshot and content so the AI can read and discuss it.
Last updated: July 6, 2026
The Web Node pulls a live webpage into your canvas. It captures a screenshot, parses the page into readable content, and makes both available to connected chats.
Adding a webpage
- Double-click the canvas and choose the web icon. The empty node prompts: "Enter a webpage URL to capture a screenshot and parse its content…"
- Paste or type the URL into the input
- Click Go
While the page loads, the button shows Scraping.... If the capture fails, red error text appears — you can correct the URL and try again.
What you get
Once scraping finishes, the node shows:
- A screenshot of the page, or the parsed markdown content
- The site's domain in the node footer
- A full-content dialog — open it to read everything that was extracted
Using it as context
Connect the web node to a chat node (drag an edge, or @-mention it from the chat). The parsed page content flows in as context, so you can ask questions about the article, compare it with other sources, or quote from it. See Connections.
Paywalled or heavily JavaScript-driven sites may fail to parse or return incomplete content. Retry, or use a different source (for example an archived or reader-friendly version of the page).
Use cases
- Research — capture several articles on a topic, connect them all to one chat, and ask for a synthesis with the sources side by side.
- Writing — pull in a reference piece and ask a connected chat to match its structure or argue against it.
- Studying — save documentation pages or tutorials and ask the AI to explain the parts you're stuck on.
- Content creation — capture a competitor's landing page and brainstorm how your version should differ.
- Planning — collect pricing pages, reviews, or specs for options you're comparing, and let the AI build the comparison.