n8n-nodes-reader-view

n8n community node that extracts the readable article (title, byline, main content, excerpt, etc.) from a URL or HTML, using Mozilla Readability — the same engine behind Firefox Reader View.

Useful for RSS pipelines, AI summarisation, newsletter digests, and any workflow where you want clean article text instead of raw HTML.

Install

Self-hosted n8n Settings → Community Nodes → Install → n8n-nodes-reader-view.

n8n Cloud Community nodes on Cloud must go through n8n's verification process. See Publish for verification. Once published to npm and submitted, verified users can install it from the Cloud node panel.

Node: Readability

Input Source — one of:

• URL — the node fetches the page over HTTP and parses it
• HTML String — pass raw HTML in as a string (from HTTP Request, Webhook, etc.)
• Binary HTML — parse an HTML file from a previous node's binary data

Options

| Option | Default | Description | | --- | --- | --- | | Char Threshold | 500 | Minimum length of the extracted article | | Keep Classes | false | Preserve class attributes on the output HTML | | Max Elements to Parse | 0 (no limit) | Hard cap on elements Readability processes | | Number of Top Candidates | 5 | How many candidate roots Readability ranks | | Probably Readable Only | false | Skip pages Readability thinks are unlikely articles | | Remove Links | keep | Post-process anchors: keep leaves them, unwrap removes <a> but keeps its text/children, strip removes the anchor and its content | | Unwrap Image Tables | false | Replace any <table> containing exactly one <img> with just the image. Fixes Substack and similar newsletter emails that wrap images in layout tables. | | Videos | keep | How to handle <video>, known-host <iframe>s, and newsletter video preview images. remove deletes them; qr replaces them with an inline SVG QR code of the video URL (handy for Kindle). | | Debug | false | Readability debug logging to stderr | | Request Timeout (ms) | 15000 | Only used for URL source | | User Agent | Mozilla/5.0 … | Only used for URL source |

Output (one item per input, even for non-readable pages):

{
  "readable": true,
  "url": "https://example.com/article",
  "title": "…",
  "byline": "…",
  "dir": "ltr",
  "lang": "en",
  "content": "<div>…cleaned HTML…</div>",
  "textContent": "…plain text…",
  "length": 12345,
  "excerpt": "…",
  "siteName": "…",
  "publishedTime": "2026-01-01T00:00:00Z"
}

If the page cannot be parsed as an article, the output is { "readable": false, "url": "…" }.

Develop

npm install
npm run dev    # runs tsc --watch inside a throwaway n8n
npm run build
npm run lint

License

MIT