n8n community node

Convert, scrape and crawl inside your workflow.

The official EnConvert node for n8n: six resources, sixteen operations, one credential. It reads the file formats n8n cannot, renders pages into markdown and real binary files, and turns a whole site into passages you can embed, all without an HTTP Request node in sight.

@enconvert/n8n-nodes-enconvert

Install this package name from Settings, Community nodes.

Why this node

Fills exactly what the built-in file node skips.

n8n's Extract From File reads CSV, HTML, JSON, ICS, ODS, PDF text, RTF, TXT, XLS and XLSX. It does not read Word, PowerPoint or EPUB, it cannot render a web page, and on n8n Cloud you cannot install the command-line tools that would fill those gaps.

You needThe node gives you

Word, PowerPoint, Excel, EPUB or RTF as text

File, Convert to Markdown. The text lands in the item, ready for the next AI node.

Almost anything as a PDF

File, Convert to PDF, with page size, orientation, margins, header and footer.

A web page as clean markdown

Web Page, Scrape. Roughly six times smaller than raw HTML, which cuts the token cost downstream.

A screenshot or page PDF as a real file

Web Page, Scrape returns them as n8n binary, ready to email or upload.

A whole site in a vector store

Website, Crawl. One item per passage, each carrying its source URL, title and heading path.

Resources

Six resources, sixteen operations, one node.

Every operation runs against the same EnConvert API, so your plan allowance, rate limits and features apply exactly as they do everywhere else.

  1. File

    Convert documents, spreadsheets, presentations, web pages, images and text into PDF; pull Word, PowerPoint, Excel, PDF, EPUB, RTF and HTML into markdown; convert between JSON, CSV, XML, YAML, TOML and HTML; or split up to 200 files per run into overlapping passages ready to embed.

    • Convert to PDF
    • Convert to Markdown
    • Convert Data
    • Split Into Chunks

  2. Image

    Convert between JPEG, PNG, WebP, HEIC and SVG, or turn a PDF page into JPEG. Compress shrinks a PNG, JPEG or WebP towards a target size, returning the smallest result achieved rather than failing on an unreachable target.

    • Convert Format
    • Compress

  3. Web Page

    Scrape one page into markdown, cleaned or raw HTML, links, images and structured data, with screenshots and PDFs as binary and about an hour of caching. Scrape Many renders up to 1000 pages in one run. Extract pulls named fields using plain language, a field list or a JSON Schema.

    • Scrape
    • Scrape Many
    • Extract

  4. Website

    Map lists a site addresses without rendering them, which is fast and uses none of your page allowance. Crawl reads the whole site and returns passages ready to embed, each carrying its source URL, title and heading path.

    • Map
    • Crawl

  5. Search

    Search the web, news, images, scholar, patents or maps, one item per result. Scrape Top Results renders the top few and attaches their markdown, so one run gives you the results page and the page contents.

    • Search

  6. Job

    Get looks up any job by ID and picks the right endpoint from the prefix, so scrape, multi-page scrape, crawl and conversion IDs all work. Get Many lists recent crawls, Cancel stops a running one, and Retry Webhook resends a completion callback.

    • Get
    • Get Many
    • Cancel
    • Retry Webhook
Setup

One private key, one credential.

The credential test calls GET /v1/whoami, which only accepts private keys, so a wrong key fails the moment you paste it rather than on your first real run.

Generate a private key

Open Dashboard, API keys and create a private key. It starts with sk_ and is shown once.

Add the credential

In n8n, open the EnConvert node, choose Create new credential, and pick EnConvert API.

Paste and test

Paste the key, save, then press Test. A green result means the key is live and your plan is active.

Leave the base URL alone

It defaults to api.enconvert.com. Change it only if you were given a different endpoint.

Public keys that start with pk_ are for browser widgets and are rejected by every operation in this node.

In practice

Four workflows, start to finish.

Real node chains from the docs. Each one replaces a small pile of HTTP Request, Code and loop nodes.

A Word document into a summary

  1. Google Drive (Download)
  2. EnConvert: File, Convert to Markdown
  3. Basic LLM Chain

The markdown lands at {{ $json.text }}, so the LLM node reads it with no glue in between.

Invoice HTML into a PDF attachment

  1. Set (build HTML)
  2. Convert to File
  3. EnConvert: File, Convert to PDF
  4. Gmail (Send)

Build the HTML, set A4 and margins, and the PDF arrives as binary ready to attach.

A website into a vector store

  1. EnConvert: Website, Crawl
  2. Embeddings
  3. Pinecone

Crawl from the sitemap, cap the page count, and every item carries content, sourceUrl, title and headingsPath.

A daily screenshot into Slack

  1. Schedule Trigger
  2. EnConvert: Web Page, Scrape
  3. Slack (Upload file)

The PNG arrives as binary under screenshot, so the Slack node uploads it directly.

Questions about the n8n node

The community node package is @enconvert/n8n-nodes-enconvert. Install it from Settings, Community nodes, Install in the n8n editor, or run npm install in a self-hosted instance data directory.
n8n Cloud only installs community nodes that have passed n8n verification review. Until then, install it on a self-hosted instance, or call the API directly with the HTTP Request node, which works on every plan.
A private API key that starts with sk_, created in Dashboard, API keys, and added as an EnConvert API credential. Public keys that start with pk_ are rejected. The Test button calls GET /v1/whoami to confirm the key works before you build anything on it.
Extract From File reads CSV, HTML, JSON, ICS, ODS, PDF text, RTF, TXT, XLS and XLSX. It does not read Word, PowerPoint or EPUB and it cannot render a web page. This node covers exactly those gaps, plus the reverse direction: almost any file into PDF.
Every File and Image operation offers three response formats: File, Text, or URL Only. Use URL Only above 16 MB, which is n8n default maximum item size and is not adjustable on Cloud. Download links expire after 15 minutes. Scrape Many, Crawl and Split Into Chunks run as jobs with a 120 second default wait; on n8n Cloud keep Max Wait Time under 300 seconds.
Yes, every operation is exposed as an agent tool. Web Page Scrape, Search, Web Page Extract and Website Map are the four worth exposing, because an agent can supply a URL or a query but cannot hand over a file.

Add the node to your workflow.

Sixteen operations, one credential, MIT licensed and free to install.

Read the n8n guideView on GitHub