n8n-nodes-pline

An n8n community node for the Pline API. Pline fetches web
content for you and handles blocked pages, retries and access escalation server-side: scrape one
page, batch-scrape a list of URLs, crawl a whole site, map a site's URLs, or get parsed web search
results.

Pline is a paid service with a free tier. You bring your own API key; this node never includes or
falls back to a shared key, and every request counts against your own workspace's credits.

• Get a Pline API key
• Set up the credential in n8n
• Installation
• Operations
• Example workflow
• Errors
• Compatibility
• Development
• Resources

Get a Pline API key

1. Sign up at https://app.pline.ai/signup. The free tier includes credits to try the API.
2. Sign in at https://app.pline.ai/login and open API keys. A key is created for you when
   you sign up; copy it. Keys can be rotated and revoked there at any time.
3. Keep the key private. n8n stores it encrypted; never paste it into workflow fields or
   expressions.

Every request is billed in credits by what it needed (a plain fetch, a browser, a stronger proxy,
AI extraction). See Credits and usage.

Set up the credential in n8n

1. In n8n open Credentials, click Add credential, and search for Pline API.
2. Paste your API key into API Key (the field is masked).
3. Click Save. n8n tests the key with a request that spends no credits: a status lookup for a
   job that cannot exist. A valid key answers "not found", which the test reports as success; a
   wrong or revoked key answers 401 or 403 and the test shows which.

Every operation of the Pline node uses this credential and sends the key as the x-api-key
header to https://apix.pline.ai/v1. Nothing else is configurable.

Installation

Follow the community nodes installation guide
and install n8n-nodes-pline. On n8n Cloud, verified community nodes can be installed from the
nodes panel.

Operations

All optional fields are sent only when you set them, so the API's own defaults apply otherwise.

Scrape: Scrape a Page

POST /scrape. Fetches one URL synchronously and returns a flat object: data (one key per
requested format: html_body, clean_html, links, markdown, clean_markdown, json,
screenshot, screenshot_full_page), the page's status, final_url, session_id,
request_id, effective_geolocation and site_tier.

Screenshot outputs are temporary signed URLs; download them promptly. A 200 whose data is
empty is a soft failure: retry with Require a Browser, or without a Basic proxy pin.

Batch Scrape

POST /batch/scrape and friends. The same per-page options applied to a list of URLs, as a
background job that survives restarts.

Start takes URLs one per line or comma-separated (an expression may return an array), the output
formats, the scrape options above, Maximum URLs, and delivery options: Email, Webhook URL,
Webhook Events (started, completed, failed), Webhook Headers and Webhook Metadata.

Crawl

POST /crawl and friends. Discovers pages by following links from a seed URL and scrapes them,
as a background job.

Start takes the URL, Maximum Pages (default 100 in the node; the API allows up to 10000), Output
Formats, Crawl Options (Include Paths and Exclude Paths as regexes, Maximum Discovery Depth,
Sitemap, Ignore Query Parameters, Ignore Robots.txt, Crawl Entire Domain, Allow Subdomains,
Email, Webhook fields, timeouts) and Page Options, the per-page fetch controls applied to every
page. Each crawled page is billed at the scrape cost for the options you choose, so size a site
with Map first and keep Maximum Pages tight.

Crawl results are ZIP files; use the HTTP Request node to download the URLs from s3PresignedUrls
and the Compression node to unpack them.

Waiting for a job

Get Status and Get Results have a Wait for Completion toggle. When on, the node polls every 5
seconds until the job is completed, failed or cancelled, or until Maximum Wait (default 120 s,
at most 600 s) passes, then returns the latest job object. A job that is still running is not an
error: check status. For long jobs, prefer the Email or Webhook delivery options and an n8n
Webhook trigger over polling.

Map: Map a Site

POST /map. Synchronous URL discovery from the site's sitemaps, with search as a fallback. Returns
one item per link (url, type, title, description) or, with Split Into Items off, the whole
response. Options: Sitemap, Include Subdomains, Ignore Query Parameters, Ignore Cache, Maximum
Links, Timeout.

Search: Search

GET /serp. Parsed web search results. Returns one item per organic result (position,
title, link, snippet, domain, page), or with Split Into Items off the whole response,
which in Full mode also carries knowledgeGraph, peopleAlsoAsk, relatedSearches and
totalOrganicResults. Options: Country, Language, Number of Results, Time Range, Device, Mode
(Fast is cheaper than Full), Page Content (also fetch each result page; one extra credit per page)
and Only Main Content.

Example workflow

Scrape a pricing page as Markdown and extract the plans as JSON:

1. Manual Trigger.
2. Pline node: Resource Scrape, Operation Scrape a Page, URL https://example.com/pricing,
   Output Formats Markdown and JSON, Options → Prompt List every plan with its name, monthly price and included credits.
3. The output item has data.markdown and data.json. Connect a Google Sheets or Slack
   node and reference {{ $json.data.json }}.

Batch example: a Pline Map a Site node feeds its items into a Pline Batch Scrape →
Start node with URLs set to {{ $input.all().map(i => i.json.url) }}, then a Pline Batch
Scrape → Get Results node with Job ID {{ $json.id }} and Wait for Completion on returns one
item per page.

Importable JSON for the first example:

{
  "nodes": [
    { "parameters": {}, "name": "Manual Trigger", "type": "n8n-nodes-base.manualTrigger", "typeVersion": 1, "position": [0, 0] },
    {
      "parameters": {
        "resource": "scrape",
        "operation": "scrape",
        "url": "https://example.com/pricing",
        "output": ["markdown", "json"],
        "options": { "prompt": "List every plan with its name, monthly price and included credits" }
      },
      "name": "Pline",
      "type": "n8n-nodes-pline.pline",
      "typeVersion": 1,
      "position": [220, 0],
      "credentials": { "plineApi": { "name": "Pline API" } }
    }
  ],
  "connections": { "Manual Trigger": { "main": [[{ "node": "Pline", "type": "main", "index": 0 }]] } }
}

Errors

The node shows the API's own message and code, plus what to do:

With Continue On Fail enabled, failed items are returned with an error field instead of
stopping the workflow.

Compatibility

Built and tested against n8n 1.x with n8n-workflow as a peer dependency. Requires Node.js 20.15
or newer to build. The package has no runtime dependencies.

Development

cd n8n-pline
npm install
npm run lint      # n8n community-node lint rules
npm run build     # compiles to dist/
npm test          # smoke test: asserts the exact requests each operation sends
npm run dev       # starts a local n8n at http://localhost:5678 with this node linked, hot reload

In the local n8n, add a Pline API credential with your key, add a Pline node, and run it.
The n8n CLI needs Node 22.12 or newer; on Node 22.11 prefix the commands with
NODE_OPTIONS=--experimental-require-module.

Resources

• Pline API documentation
• Pline API reference
• n8n community nodes documentation
• Issues