LinkedIn Jobs Scraper

LinkedIn Jobs Scraper

Search LinkedIn’s public guest jobs endpoint and return normalized, deduplicated job listings. The Actor supports public date, job-type, experience, and workplace filters, bounded pagination, optional detail enrichment, local HTML fixtures, and truthful run diagnostics. It does not require login credentials.

Target: LinkedIn public jobs

Use cases

• Build a current job-market dataset for a keyword and location.
• Compare public listings by recency, workplace, employment type, or experience level.
• Enrich a bounded result set with public descriptions, criteria, salary text, skills, and benefits.
• Regression-test parsing against synthetic list/detail HTML before a live run.

Input

The minimum valid input is:

{
  "keywords": "software engineer"
}

Supported fields:

Field	Type	Default / bounds	Description
keywords	string	required, 1–200 chars	Public job-search text.
location	string	United States, 1–200 chars	Location sent to LinkedIn.
maxJobs	integer	50, 1–100	Maximum unique normal records.
maxPages	integer	derived, 1–20	Optional pagination bound.
datePosted	enum	anyTime	pastMonth, pastWeek, or past24Hours are also supported.
jobType	enum	any	F full-time, P part-time, C contract, T temporary, I internship, V volunteer, O other.
experienceLevel	enum	any	LinkedIn public levels 1–6.
remoteFilter	enum	any	1 on-site, 2 remote, 3 hybrid.
scrapeJobDetails	boolean	true	Fetch public job detail HTML for each unique result.
includeDiagnostics	boolean	true	Include minimal diagnostic rows for blocked/empty/failing requests.
requestTimeoutSecs	integer	30, 10–120	Per-request timeout.
maxRequestRetries	integer	2, 0–5	Transient HTTP retry limit.
requestDelayMs	integer	250, 0–10000	Pacing between public endpoint requests.
proxyConfiguration	object	residential US proxy	Standard Apify proxy configuration; set { "useApifyProxy": false } for a direct local smoke run.

For local development, fixtureListHtml or fixtureListFiles enables deterministic list-page mode. fixtureListFile is a compatible single-file alias. fixtureDetailHtml, fixtureDetailFiles, and fixtureDetailFile provide detail pages for fixture enrichment. Fixture paths must stay inside this Actor directory and use .html or .htm.

Example with filters and bounded output:

{
  "keywords": "data engineer",
  "location": "New York",
  "maxJobs": 10,
  "datePosted": "pastWeek",
  "jobType": "F",
  "remoteFilter": "2",
  "scrapeJobDetails": false,
  "includeDiagnostics": true
}

Output

Normal dataset rows contain public business and provenance fields. jobDescription is used instead of the platform-reserved description field.

Field	Meaning
position, jobId, jobTitle, company, location, jobUrl	Stable normalized listing identity and core public fields.
companyUrl, companyLogo, isRemote, postedDate, applicantCount	Public card metadata when available.
salary, employmentType, seniorityLevel, industry, jobFunction	Public detail criteria when available.
jobDescription, descriptionBullets, benefits	Public detail content when available.
searchKeywords, searchLocation, scrapedAt	Search context and collection provenance.

The earlier detailScraped row field was intentionally removed as extraction instrumentation. Detail request/success/failure counts remain in the OUTPUT key-value summary, while business fields are retained.

Diagnostic rows, when enabled, contain only url, error, errorCode, and scrapedAt; they never masquerade as job records. The OUTPUT key-value record includes success/empty/partial status, normal and diagnostic counts, pagination, detail, deduplication, request, proxy, fixture, and bounded failure evidence.

Limits and responsible use

Requests are sequential and bounded. Responses larger than 3 MB, access challenges, and malformed public cards are handled fail-closed with summary/diagnostic evidence. LinkedIn job availability and public endpoint behavior change frequently.

Use public job data lawfully and respect LinkedIn terms, rate limits, and applicable privacy and data-protection requirements. Do not provide credentials or attempt to bypass access controls.

You can download the dataset in various formats such as JSON, HTML, CSV, or Excel.

Local development

From this Actor directory:

npm install
npm test
npm run check
apify run --purge --input-file test/inputs/fixture-basic.json
apify run --purge --input-file test/inputs/fixture-detail.json
npm run validate

The fixture runs do not contact LinkedIn. A direct public smoke input can use proxyConfiguration: { "useApifyProxy": false }; no credentials are required.