Web Page Ingestion

Memwyre includes a powerful web scraping engine that turns any URL into a clean, searchable memory — stripping away ads, navbars, and clutter automatically.

Overview

When you save a URL, Memwyre fetches the raw page, strips non-essential elements (navigation, footers, advertisements, and scripts), and extracts only the core content. Your vault stays populated with high-quality information, not noise.

How It Works

1. Save a link — Paste a URL into the Memwyre chat, add it via the browser extension, or create a new memory from your Inbox.
2. Noise reduction — Memwyre analyses the page structure and removes non-content HTML elements.
3. Content indexing — The cleaned text is processed, chunked, and embedded into your vector vault.

You can then ask questions like:

"Summarise the article I saved about React Server Components."

Tips

• Works best with article and documentation pages. Dynamic single-page apps (SPAs) that require JavaScript to render content may not extract cleanly.
• For pages behind a login or paywall, use the browser extension to capture and save content directly from your browser session instead.
• There is no limit on how many URLs you can ingest.