StepStone Jobs Scraper

StepStone runs a separate site per country, and they do not agree on the word for a job: the Dutch one lists under /vacatures/ where the German one uses /jobs/. Pick the country site, give it a keyword and a location, and the job cards come back as rows. Five sites are wired up: Germany, Austria, the UK, the Netherlands and Belgium.

StepStone sits behind DataDome, which is why plain HTTP clients get a challenge page instead of listings. This actor loads the pages the way a browser does and takes a fresh exit address whenever one gets blocked.

What you get per job

Field	Notes
jobId	StepStone's listing id. null when the URL carries no id.
title	Job title.
company	Hiring company. null when StepStone hides it on the card.
location	City or region.
salary	Only when the listing publishes one. Most German listings don't, so expect null more often than not.
postedText	As shown, e.g. vor 5 Tagen. Not a parsed date.
workFromHome	Boolean remote flag.
snippet	The description excerpt from the card, a couple of sentences.
url	Link to the listing.
site	Which country site the row came from.

Real row from a software engineer / de run:

{
  "ok": true,
  "site": "de",
  "jobId": "14295362",
  "title": "Software Engineer (m/w/d)",
  "company": "Menck GmbH",
  "location": "Kaltenkirchen",
  "salary": null,
  "postedText": "vor 5 Tagen",
  "workFromHome": false,
  "snippet": "Menck, als bedeutender Akteur der renommierten ACTEON-Gruppe…",
  "url": "https://www.stepstone.de/stellenangebote--Software-Engineer-m-w-d-…-14295362-inline.html"
}

Input

{
  "keywords": "software engineer",
  "location": "Berlin",
  "site": "de",
  "maxItems": 100
}

Field	Default	Notes
keywords	required	Job title or search terms. German works: krankenpflege, vertrieb.
location	empty	City or region. Leave it out for a nationwide search.
site	de	de, at, uk, nl, be.
maxItems	25	1 to 2000. Pages are about 25 jobs each.
proxyConfiguration	Apify datacenter	Leave it alone. The actor rotates datacenter addresses itself. Fill it in only if you own proxy servers you'd rather use.
notionConnector / notionParentId	—	Optional delivery into Notion via an Apify MCP connector.

The country site is picked by domain, not by where the request comes from, so site: "uk" gets you UK jobs regardless of the exit address. Umlauts are transliterated the way StepStone's own search does it (München → muenchen), and stepstone.nl uses its vacatures path.

Limits worth knowing

• Card data only. You get what's on the search results page. The full job description lives on the detail page and isn't fetched.
• salary is usually null, because most StepStone listings don't publish pay.
• postedText is a string in the site's language, not a timestamp. Parse it yourself if you need a date.
• Pagination stops at 50 pages or when two consecutive pages bring nothing new.
• A run has a four-minute budget. If DataDome is being difficult it stops there rather than burning compute on a search it isn't going to win.

When a search gets blocked, you don't pay

• No matches → one NO_RESULTS row, uncharged.
• Blocked on every rotation → one BLOCKED row, uncharged.
• Blocked part way through → the rows already gathered come back marked ok: false, partial: true alongside a BLOCKED row, and none of them are charged. Run it again; a different exit address usually finishes the job.

Billing

$3.00 per 1,000 jobs ($0.003 each), with no run-start fee. Flat on every plan.

You're charged once per job row with ok: true. Everything else — diagnostics, partial blocked results, duplicates — is free.