LinkedIn Jobs Scraper | Details & Companies

What you can export

Use this LinkedIn jobs scraper for role research and hiring analysis. Supply titles and locations, job/search URLs or supported company inputs to export deduplicated public guest jobs with available descriptions, salary fields and source context.

Optional company-profile reads add public company information. No LinkedIn login or uploaded CV is required; field availability and result coverage depend on the guest surfaces.

Try the example input

The Console prefill searches Software Engineer in United States and returns up to three detailed jobs. API callers should send their own search; {} exits successfully with guidance and no LinkedIn reads or result rows.

{
  "titles": [
    "Software Engineer"
  ],
  "locations": [
    "United States"
  ],
  "startUrls": [],
  "companyIds": [],
  "companyUrls": [],
  "companyNames": [],
  "mode": "detailed",
  "maxItems": 3,
  "maxRowsPerSearch": 0,
  "companyProfile": true,
  "datePosted": "anyTime",
  "easyApply": false,
  "under10Applicants": false,
  "geoId": "",
  "customUrlParams": {},
  "filters": {},
  "resumeKeywords": [],
  "splitByLocation": false,
  "splitCountry": "",
  "splitLocations": [],
  "splitTimeRanges": [],
  "splitCompanyIds": []
}

Run through the Apify API with your Actor ID:

curl -X POST "https://api.apify.com/v2/acts/YOUR_ACTOR_ID/runs?token=YOUR_APIFY_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{"titles":["Software Engineer"],"locations":["United States"],"maxItems":100}'

Read the resulting run’s default dataset as JSON, CSV or Excel. The output links also expose ERRORS, COVERAGE, SEARCH_TOTALS, METRICS and run-local STATE.

Inputs

Omitted or null values use the runtime values below. Explicit empty arrays, false and zero are preserved. Schema prefill is a Console example; it is never an API default.

Input	Type	Default	Meaning
titles	array	[]	Keywords; each title is paired with every location.
locations	array	[]	LinkedIn location hints; each is paired with every title.
startUrls	array	[]	LinkedIn job, search or company URLs; strings or objects containing url. Read together with structured searches.
companyIds	array	[]	Numeric LinkedIn organization IDs as strings.
companyUrls	array	[]	Public company root/jobs/about/posts URLs; resolve the canonical root to its numeric ID.
companyNames	array	[]	Exact company names; resolve uniquely from guest search and confirm the company root name. Ambiguity goes to ERRORS.
mode	string	"detailed"	quick reads cards; detailed adds full job descriptions and optional company profiles.
maxItems	integer	1000	Maximum unique delivered jobs across the run; 0 performs no reads. Maximum 100000.
maxRowsPerSearch	integer	0	Delivered-row cap per partition; 0 uses the 1000-card source window. Maximum 1000.
companyProfile	boolean	true	Join the public company root in detailed mode. Company identity/bound filters may still require a root read when false.
datePosted	string	"anyTime"	anyTime, past24Hours, pastWeek, pastMonth or r followed by positive seconds. Conservative observed-date post-check.
easyApply	boolean	false	Detailed only; require an explicit onsite-apply marker. Unknown is excluded.
under10Applicants	boolean	false	Detailed only; require an explicit applicant number below ten. Lower-bound captions are excluded.
geoId	string	""	LinkedIn geographic ID passed to the guest query.
distance	integer	null	Optional nonnegative distance hint in the guest query; units and nearby results are controlled by LinkedIn.
customUrlParams	object	{}	Supported guest parameters only: keywords, location, geoId, distance, f_TPR, f_C, f_AL, f_EA, sortBy; legacy f_WT/f_JT/f_E are translated to exact local filters.
filters	object	{}	Exact filters and dynamic flags described below; filters alone start a broad search.
resumeKeywords	array	[]	Local literal keyword/alias matching in full descriptions; no uploaded CV, model calls or external enrichment.
splitByLocation	boolean	false	Add capped partitions for distinct visible card locations; derived city rows must match the full observed location string.
splitCountry	string	""	With splitByLocation, replace each initial search location with this country hint; preserve derived city partitions.
splitLocations	array	[]	Additional location partitions preserving each initial query’s other constraints.
splitTimeRanges	array	[]	Additional time partitions, each r followed by positive seconds; preserve other constraints.
splitCompanyIds	array	[]	Additional numeric-company partitions preserving other constraints.

Exact filters

filters.workplaceTypes accepts Remote, Hybrid, On-site; employmentTypes accepts Full-time, Part-time, Contract, Temporary, Internship, Volunteer, Other; experienceLevels accepts Internship, Entry level, Associate, Mid-Senior level, Director, Executive. These match explicit observed values, case-insensitively. Unknown values never pass. Generic mentions of remote benefits are not workplace proof. industries and jobFunctions require any requested phrase’s words in the corresponding criterion. easyApply and under10Applicants are equally strict. An applicant caption like “Be among the first 25 applicants” is a range, and cannot pass an under-ten filter.

titleInclude is an OR of phrases, each requiring all its words in any order. titleExclude rejects any such phrase. skipJobIds excludes exact IDs; locationExclude excludes case-insensitive substrings; companyExclude excludes exact case-insensitive employer names. All are string arrays. Quick mode supports these card-level filters and observed dates; detail-level filters and skill matching require detailed mode.

Legacy search-URL f_WT, f_JT and f_E codes are translated into these local labels, rather than trusting LinkedIn to enforce them. URL and global label constraints must overlap; derived partitions retain those constraints. Global date, easy-apply and under-ten requirements also apply to URL sources and individual jobs. A URL date window and global date window use the narrower requirement. These modifiers do not add an unrelated broad search when source inputs are present. Supported custom parameters configure structured searches; unknown parameters and filters are declined into ERRORS. LinkedIn keyword/location/distance/geoId inputs are search hints, not promises of exact text or geography. Supply title filters when exact role wording matters. Automatic observed-city partitions additionally require an exact full location-string match. Explicit location partitions retain LinkedIn’s nearby-result behavior.

Dynamic filters and skill matching

These filters set dynamicFilterMatch and retain both matching and nonmatching jobs: jobFunctionInclude/jobFunctionExclude, jobIndustryInclude/jobIndustryExclude, companyOrganizationTypeInclude/companyOrganizationTypeExclude (string arrays), companySizeMin/companySizeMax, companyFoundedDateMin/companyFoundedDateMax, companyFollowersCountMin/companyFollowersCountMax (nonnegative integers), and excludeRecruitingAgencies, requireRecruiterProfile, requireSalaryInfo (booleans). Include takes precedence over exclude for each pair. Unknown company facts pass that dynamic check; they are not silently guessed. Company size bounds apply to the exact observed employee count, never a size band.

{"titles":["Engineer"],"locations":["Canada"],"maxItems":25,"filters":{"workplaceTypes":["Hybrid"],"titleExclude":["Sales Engineer"],"requireSalaryInfo":true},"resumeKeywords":[{"keyword":"C++","aliases":["CPP"]},{"keyword":"JavaScript","aliases":["JS"]}]}

Skill scoring checks literal keywords or aliases against description text with word boundaries. JS does not match JSON. This is a transparent text match, not an assessment of employability, semantic ranking or resume processing.

Source modes and partitions

{"startUrls":["https://www.linkedin.com/jobs/search/?keywords=Data%20Engineer&location=Canada",{"url":"https://www.linkedin.com/jobs/view/4419969671"},"https://www.linkedin.com/company/microsoft/jobs/"],"companyIds":["1035"],"companyNames":["Microsoft"],"maxItems":20}

Company URLs resolve their public root to a numeric organization ID. Company names must resolve uniquely to an exact name in a guest search and the company root; unresolved names go to ERRORS and never broaden into all jobs. A job URL must expose its own matching ID and a full description. Search cards may carry their own job link or a child full link; either form must match the card’s numeric job ID. Quick search mode reads cards without job-detail requests; company-filter identity checks may read a company root. Individual job URLs always require a detail read, including in quick mode.

{"titles":["Engineer"],"locations":["United States"],"mode":"quick","maxItems":10000,"maxRowsPerSearch":1000,"splitLocations":["Canada","United Kingdom","Germany","France","India","Singapore","United Arab Emirates","Japan","Brazil","Australia"]}

Every search stops before the measured start=1000 boundary. Ten or more populated capped partitions can supply 10,000 unique jobs; this is a run capacity, not guaranteed yield. Explicit location, time and company partitions preserve other query constraints and are added once to each initial search; time/company splits can overlap. Automatic location splitting discovers locations from visible cards; it is not a complete country or city inventory. All partitions deduplicate by job ID, with first-delivered query context retained. An initial structured query equivalent to a search URL shares one partition and uses structured context (inputUrl is empty), including after restart. COVERAGE states whether a partition ended on an empty page or hit a source/row/global/charge limit. SEARCH_TOTALS preserves the displayed total separately; it is not an expected final dataset size.

On an Apify migration/restart, the existing run dataset supplies delivered IDs and counts. STATE retains the pending partition queue and skips completed partitions; incomplete, uncapped partitions replay from offset zero and deduplicate delivered IDs. Row-capped partitions report row_cap without further target reads. Replayed coverage/totals replace their earlier partition record; earlier errors remain available. A partial detail, company or totals failure prevents a complete coverage claim. Resume is within one run, not between new runs or a change-feed mode. No error becomes a dataset row. Invalid/unresolved inputs and source failures are recorded in ERRORS; mixed-source runs can still deliver valid jobs. A run where every source attempt fails reports FAILED. A valid empty search/filter result exits successfully. maxItems: 0 makes no target requests. Native PPE charge limits stop delivery and further source reads; at most an already-finished batch of eight enrichments precedes a binding push.

Output

Every job has the same nullable shape in quick and detailed modes. Quick fields from cards remain populated; unavailable detail/company facts are null or empty arrays. Full descriptions, salary, poster, external apply and workplace fields depend on the public page. No missing fact is inferred. Applicant numbers can be lower bounds; source captions remain available. applyUrl is the public LinkedIn job landing page; externalApplyUrl is nullable when the guest route exposes no usable offsite target. Posting dates retain their observed precision; exact timestamps and expiry are not manufactured.

Field	Meaning
id	LinkedIn job ID; deduplication and billing identity.
jobUrl	Canonical public job URL.
title	Observed job title.
companyName	Observed employer; verified root name when joined.
companyUrl	Canonical company root URL; null for absent or other employer-link families.
companyId	Verified organization ID, nullable without a join.
companyLogo	Public company/card logo URL.
location	Displayed location; not a geocoded address.
locationParsed	Nullable; no guessed geocoding.
postedTime	Observed relative posting-age caption.
publishedAt	Card time datetime, preserving source precision.
postedAtTimestamp	Nullable; no invented millisecond posting timestamp.
benefits	Visible card benefit/badge strings.
hiringBadge	Visible card hiring badge.
isActivelyHiring	True only from a visible actively-hiring badge; otherwise null.
isEarlyApplicant	True only from an explicit early-applicant badge; otherwise null.
description	Full public description as plain text.
descriptionHtml	Inner HTML of the full public description.
applicantText	Original applicant caption, including lower-bound qualifiers.
applicantsCount	Number in the caption; may be a lower bound, not an exact count.
applyUrl	Canonical LinkedIn job landing URL in both modes.
externalApplyUrl	Offsite application target exposed by an apply anchor or embedded applyUrl JSON comment; null for Easy Apply or when no usable offsite target is exposed.
applyType	EASY_APPLY for an onsite control; EXTERNAL with a usable offsite target; NOT_EXPOSED when details expose no usable offsite target; null when quick mode skips details.
easyApply	True for an onsite control; false for a usable offsite target; otherwise null.
seniorityLevel	Labeled job criterion.
employmentType	Labeled job criterion.
jobFunction	Labeled job criterion.
industries	Labeled job criterion.
workplaceType	Remote/Hybrid/On-site from an explicit role statement or title suffix; otherwise null.
workRemoteAllowed	True only for verified Remote; false for Hybrid/On-site; unknown null.
salary	Visible compensation text, otherwise null.
salaryInfo	Observed currency-and-amount strings.
salaryMin	Parsed visible minimum; no currency conversion.
salaryMax	Parsed visible maximum.
salaryCurrency	Explicit currency, otherwise null; bare $ is ambiguous.
salaryPeriod	Explicit hour/day/week/month/year period, otherwise null.
salarySource	Nullable; no guessed source.
salaryInsights	Nullable; no hidden insights fetched.
expireAt	Nullable; not exposed by the measured guest route.
jobState	CLOSED only when the source explicitly says no longer accepting; otherwise null.
repostedJob	True only when explicitly labeled reposted; otherwise null.
verified	Nullable; no guessed verification.
insights	Empty when unavailable.
posterFullName	Public recruiter card name, otherwise null.
posterProfileUrl	Public recruiter card link; its profile is never fetched.
posterHeadline	Public recruiter card headline.
posterPhoto	Public recruiter card image.
contactEmail	An email literally present in the job description; no contact enrichment.
yearsOfExperience	Literal English years-of-experience phrases with context.
companyDescription	Public Organization JSON-LD description.
companyWebsite	Labeled company website; never visited.
companyEmployeeCount	Exact public Organization numberOfEmployees, otherwise null.
companySize	Labeled size band, distinct from the exact count.
companyFollowerCount	Public follower count.
companyIndustries	Labeled company industry.
companyTagline	Public Organization JSON-LD slogan; null when absent. Distinct from industry and company description.
companyHeadquarters	Public Organization address object, otherwise null.
companyHeadquartersText	Labeled headquarters text.
companyType	Labeled organization type.
companyFoundedYear	Labeled founded year.
companySpecialties	Visible specialties.
companyOfficeLocations	Visible office-location strings, empty if absent.
companyAffiliatedPages	Visible affiliated-page names, empty if absent.
companyRecentPosts	Visible public post text, URL and source datePublished; dates may be incomplete.
companySourceUrl	Verified company root read URL.
companyScrapedAt	Company observation UTC.
dynamicFilterMatch	Combined dynamic-filter flag; does not remove rows.
matchedKeywords	Matched requested skill keywords.
unmatchedKeywords	Unmatched requested skill keywords.
keywordMatchScorePercentage	Matched keyword fraction × 100, rounded; null when no keywords.
scrapingInfo	Input URL, title, location, query, offset, page, index and partition key.
sourceUrl	Exact guest list/detail URL used for the row.
scrapedAt	Row observation UTC.

Access, limits and billing

The Actor uses ordinary anonymous HTTP on LinkedIn public guest search, detail and company roots. It uses Apify’s automatically allocated US datacenter proxy pool and up to eight concurrent reads. Company reads may reuse a successful sticky session for less than 50 seconds; classified blocks retire it and retry with fresh sessions, up to ten attempts with backoff capped at two seconds. Other reads use fresh sessions and at most three classified attempts with shared backoff for blocking/rate limits/transient failures. Redirects, unknown pages and absent identities are errors. Login-gated profile data, company employees, website crawling and residential fallback are outside this version.

One successfully delivered unique job is one native dataset-item billing event. Errors, totals, coverage and state are key-value-store records. See the Store Pricing tab for current pay-per-event prices and any Actor Start charge. Compute and proxy costs still depend on filters, retries and company diversity. METRICS response-body bytes are useful diagnostics, not an invoice or exact billable network traffic.

LinkedIn can change its guest surfaces, omit fields, cap results or block requests. The public route offers no exhaustive coverage or long-term reliability guarantee.

Guest pages may expose only a sign-in or join modal for an offsite application. Those account controls are not application destinations: externalApplyUrl stays null and applyType is NOT_EXPOSED. The Actor does not sign in or derive an application URL from a company website.