XING Jobs Scraper — XING Germany, Austria & Switzerland Jobs
Pricing
from $0.36 / 1,000 jobs
XING jobs scraper for Germany, Austria and Switzerland: title, company, city, employment type, XING salary estimate, posting and expiry dates, full description — plus job alerts: only jobs new since your last run. Search by keyword and location or paste a XING search URL. No login.
One search per line, the same words you would type on XING — e.g. "python". Many searches in one run cost one start fee.
Or paste search result URLs from XING — set any filters on the site first, then copy the address, e.g. https://www.xing.com/jobs/search?keywords=python&location=Berlin. Each URL is its own search.
Same as the "where" box on XING, e.g. "Berlin". Leave empty for the whole country.
Only jobs posted in the last N days. Leave empty for all.
Schedule this Actor daily and pick "Only new" to get a cheap new-jobs feed. The first run returns everything (there is nothing to compare with yet).
Adds one free row with type "summary" per search: status (ok / no-results / all-filtered / nothing-new / error), listed, read, returned, new, closed. Off by default so your job table stays clean; the same summary is always in the run status message and the OUTPUT record.
Comma-separated. Keep jobs whose title contains any of these, e.g. "senior, lead".
Comma-separated. Filters the returned jobs by their location text.
Keep only these levels, read from the job title (the same words on every northbell job scraper; each row also gets a seniority column). Product, account and other non-people "managers" count as mid level.
Normalised from the site's own words (Vollzeit, CDI, Jornada completa, workload %, …) into the same five types on every site; each row also gets an employmentTypes column. Roles whose type is not stated are left out when you set this, and counted in OUTPUT.
Keep only jobs whose salary could be read as a number.
Read from the job text, not from the language the ad is written in: "German C1 required" in an English ad still counts as required. Jobs whose text says nothing about it are left out of both filters (unknown is not "not required"). Every row has localLanguageNeed, localLanguageLevel and the sentence it was read from, whichever you choose.
The complete job text as plain text. One extra request per job returned (only the jobs you keep after filtering). Turn off for faster runs.
Remembers which jobs each search returned, so the next run can mark new ones (and closed ones, when the whole search fits in Max jobs).
