LinkedIn Jobs Scraper - Full Descriptions + Company, No Login

Search LinkedIn's public job listings by keyword and location and get every posting back with its full description, company name and company page, location, posting date, applicant count, seniority level, employment type and industries, plus salary where LinkedIn shows one. No LinkedIn login, no cookies, no account of your own. $1.49 per 1,000 complete job postings, no start fee. Postings LinkedIn would not open for us are still written, marked partial, and not charged. See Pricing.

Input:

{
  "keywords": ["software engineer", "data analyst"],
  "location": "United States",
  "datePosted": "week",
  "jobType": ["fulltime"],
  "experienceLevel": ["mid-senior"],
  "includeDescription": true,
  "maxItems": 200
}

Output - one row per posting (recordType: "job"); per-term status rows and the run summary are free. Sample from a real run, description shortened:

{
  "recordType": "job",
  "jobId": "4479196690",
  "url": "https://www.linkedin.com/jobs/view/4479196690/",
  "title": "Software Engineer I, Fullstack (Search)",
  "companyName": "Affirm",
  "companyUrl": "https://www.linkedin.com/company/affirm",
  "location": "Madison, WI",
  "postedAt": "2026-10-16",
  "publishedAt": "2026-10-16",
  "postedAtText": "1 week ago",
  "applicantsText": "41 applicants",
  "benefitsText": "Actively Hiring",
  "seniorityLevel": "Mid-Senior level",
  "employmentType": "Volunteer",
  "jobFunction": null,
  "industries": "Financial Services",
  "salaryText": null,
  "descriptionText": "At Affirm, we exist for the moments that matter—giving people a clear, predictable way to pay over time…",
  "applyMethod": "external",
  "applyUrl": null,
  "query": "software engineer",
  "searchLocation": "United States",
  "rank": 3,
  "partial": false,
  "partialReason": null,
  "collectedAt": "2026-10-24T09:43:15.750Z"
}

Values are what LinkedIn shows on the posting. In the sample above the employer itself set the employment type to "Volunteer"; the Actor does not correct it.

Limits:

• Public data only. The guest job search and the guest job page that LinkedIn serves to logged-out visitors. Nothing behind sign-in: no "who's hiring" names, no recruiter or poster profiles, no Easy Apply forms, no "see how you compare" insights.
• applyUrl is usually null. LinkedIn shows the employer's apply link to logged-out visitors on only some postings; elsewhere it sits behind sign-in. applyMethod still says whether the posting is applied to on the employer's site (external) or on LinkedIn (linkedin).
• No on-site / remote / hybrid filter. LinkedIn neither applies that filter nor shows a posting's workplace type to logged-out visitors, so workplaceType is refused with an error rather than silently ignored. Put "remote" in your location or search terms instead.
• Job type and experience level are filtered by the Actor, not by LinkedIn. LinkedIn ignores those two filters for logged-out visitors, so the Actor opens each posting and keeps it only if LinkedIn's own Employment type / Seniority level on the page matches. Dropped postings are free, but a narrow filter makes a run slower. Many employers set seniority to "Not Applicable", which matches no level. datePosted is applied by LinkedIn's search itself.
• At most 1,000 postings per search. That is LinkedIn's own ceiling for one keyword + location + filter combination. Use more terms, locations or filters for more.
• Fields LinkedIn leaves out are null, never guessed. salaryText is filled only where LinkedIn shows a salary field (about one posting in ten in our tests; a range written inside the description stays in descriptionText). jobFunction is currently not shown to logged-out visitors at all, so it is null on every row; the field is kept so your schema does not change if LinkedIn shows it again.

What it does

• Searches each term separately, in the location you give, filtered by date posted, job type and experience level.
• Opens each posting's public page for the full description (as plain text in descriptionText and as LinkedIn's HTML in descriptionHtml), the applicant count and the job criteria. Turn includeDescription off for a faster, search-card-only list.
• Deduplicates across the whole run: a posting that matches several of your terms is written once and charged once.
• Keeps LinkedIn's posting ID (jobId), LinkedIn's posting date (postedAt / publishedAt, a calendar date, plus postedAtText as shown) and our collection time (collectedAt) as separate fields.
• Says what happened per search term in a free query_status row: how many postings it returned, how many repeats were skipped, how many pages were read and why paging ended. A term LinkedIn answered with nothing is an empty result; a term cut short by a limit, refusal or error is partial: true with the reason. The two are never confused.
• Retries every request LinkedIn refuses (HTTP 429) from a different IP. A posting whose page still cannot be read, or that was closed after it was listed (HTTP 404, partialReason: "job_closed"), is written from its search card, marked partial and not charged.
• Caps itself: maxUpstreamRequests is a hard ceiling on requests to LinkedIn for the whole run, retries included, derived from your input when you leave it empty, so a run is never unbounded.

Input

Field	What it does
keywords	Search terms, one per line. Required.
location	City, region or country as you would type it into LinkedIn, e.g. United States, London, England, United Kingdom, Bengaluru. Empty = worldwide.
datePosted	any, 24h, week or month.
jobType	Any of fulltime, parttime, contract, temporary, internship, volunteer, other. Matched against the posting's own Employment type.
experienceLevel	Any of internship, entry, associate, mid-senior, director, executive. Matched against the posting's own Seniority level.
includeDescription	Default true: open each posting for its description and criteria.
maxItems	Maximum job rows for the whole run (default 100), filled term by term in the order given.
maxUpstreamRequests	Optional hard ceiling on requests to LinkedIn.

Output

• recordType: "job" - one per posting. Charged only when partial is false.
• recordType: "query_status" - one per search term. Free.
• recordType: "RUN_SUMMARY" - the last row, also stored under the key RUN_SUMMARY: postings collected and charged, partial postings by reason, duplicates, empty and incomplete terms, requests and retries. Free.

partialReason values: blocked (LinkedIn kept refusing after every retry), job_closed (posting removed after it was listed), upstream_request_cap, charge_limit_reached, upstream_error, not_collected (the run stopped before this term).

Pricing

Pay per result: $1.49 per 1,000 complete job postings, no start fee. Partial rows, status rows and the run summary are free. Set a maximum charge per run in Apify and the Actor stops at it, keeping what it already collected and marking the term charge_limit_reached.

Is it legal?

It reads only pages LinkedIn publishes to logged-out visitors, and it collects no personal profiles: no recruiter, poster or employee names. Company names and company pages are public business information. You are responsible for how you use the data; check LinkedIn's terms and the data-protection rules that apply to you.