LinkedIn Jobs Scraper — exact posting dates, no login

Scrape public LinkedIn job postings into typed rows. Every posting carries the exact timestamp LinkedIn published it with, not a parsed "6 days ago" string, and the run reports the fill rate of every field so you know the data is complete before you use it.

Scope: job listings only. This actor reads company-published postings from LinkedIn's public job search. It does not touch member profiles, contact details, or anything behind a login.

What you get per posting

Field	Example	Notes
jobId	4437461678	LinkedIn's own id, use it to deduplicate across runs
title	Software Engineer, Early Career
company	Notion
companyUrl		LinkedIn company page
companyLogoUrl
location	San Francisco, CA
postedDate	2026-07-28T00:00:00.000Z	Exact timestamp from LinkedIn, not a relative string
postedDaysAgo	6	Derived from the timestamp
isNewPosting	true	LinkedIn's own "new" marker
benefitBadge	Be an early applicant	LinkedIn's own badge when present
jobUrl		Direct link to the posting
searchKeyword, searchLocation, pageNumber, scrapedAt		Provenance of every row

With Include full job description enabled, each row also carries:

Field	Notes
description	Full posting text
seniorityLevel, employmentType, jobFunction, industries	LinkedIn's own criteria list
applicantCountText	Applicant count when LinkedIn shows it

Input

{
  "keywords": ["software engineer", "data engineer"],
  "location": "United States",
  "datePosted": "last_week",
  "workplaceType": "remote",
  "experienceLevel": "mid_senior",
  "maxJobs": 1000
}

Filters: date posted, job type, experience level, workplace type (on-site / remote / hybrid), sorting by relevance or date, plus an optional geoId when you need a more precise location than the text field allows.

Why this one is cheap to run

LinkedIn's public job search returns server-rendered HTML, so this actor needs no browser at all. That keeps the cost per thousand rows far below a browser-based scraper — which is why the price per result is what it is.

Enabling descriptions adds one request per posting, so it multiplies run cost. It is off by default for that reason.

The result cap is real

maxJobs stops the run at exactly that number, and duplicates are dropped on LinkedIn's job id, so the same posting is never delivered — or billed — twice. The run also stops on its own if three consecutive pages return nothing usable.

Runs fail loudly, not silently

Every run writes a DIAGNOSTICS record with block rate, bytes per item and the fill rate of every output field:

{ "itemsExtracted": 500, "blockRate": "0%", "fieldCoverage": { "postedDate": "100%", "company": "100%" } }

If nothing was extracted, or more than half the pages were blocked, the run is marked failed with the reason instead of handing you empty columns.

Typical uses

• Tracking who is hiring for a role, and how fast postings turn over
• Feeding a niche job board or a recruiting pipeline
• Labour market research on posting volume by company, location and seniority
• Competitive hiring intelligence: which teams a company is building out

Notes and limits

• Salary is not available in LinkedIn's public job search. Measured: the search fragment contains no compensation markup at all. When an employer publishes pay, it appears inside the description body, so enable descriptions if you need it. This actor does not invent an estimate.
• The public search returns a limited depth per query. For large harvests, split the work by location, seniority or date window rather than paging very deep.
• Keep maxConcurrency moderate; the endpoint rate limits per IP.

Support

If a field stops filling or you need a filter that is not exposed, open an issue on the actor page with the input you used.