If you're building a job board, tracking who is hiring for a role, or collecting salary data, LinkedIn is the obvious source. You don't need an account for it: LinkedIn shows job postings to logged-out visitors, and that public version is what this tutorial scrapes.

I build and maintain scrapers, including a LinkedIn jobs scraper on Apify (more on that near the end, with a disclosure). Everything below was tested on 21 November 2026.

What LinkedIn shows without an account

Two pages matter.

The search results. When you scroll a job search while logged out, the page loads more results from this URL:

https://www.linkedin.com/jobs-guest/jobs/api/seeMoreJobPostings/search?keywords=python%20developer&location=United%20States&start=0 

It returns an HTML fragment with 10 job cards. start moves through the results 10 at a time, and the usual search filters work as query parameters: f_TPR=r86400 for the last 24 hours, f_TPR=r604800 for the last week, f_AL=true for Easy Apply only. Each card has the job ID (data-entity-urn="urn:li:jobPosting:<id>"), the title, the company, the location and the exact posting date.

The job page. https://www.linkedin.com/jobs/view/<id>/ is the page anyone sees when they open a job link. It has the full description, the criteria list (seniority, employment type, industries), the number of applicants and, for some jobs, a salary box.

What you won't get while logged out:

  • the external "apply on company site" link, which LinkedIn only shows to signed-in users;
  • workplace type (remote, hybrid or on-site) as a field: you have to read it from the title or the description;
  • the job function, and who posted the job.

A minimal scraper in Python

Install the two libraries:

pip install requests beautifulsoup4 

Then:

import time import requests from bs4 import BeautifulSoup SEARCH = "https://www.linkedin.com/jobs-guest/jobs/api/seeMoreJobPostings/search" HEADERS = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 " "(KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36"} def text(el): return el.get_text(" ", strip=True) if el else None def search_jobs(keywords, location, start=0): params = {"keywords": keywords, "location": location, "f_TPR": "r604800", "start": start} r = requests.get(SEARCH, params=params, headers=HEADERS, timeout=30) r.raise_for_status() soup = BeautifulSoup(r.text, "html.parser") jobs = [] for card in soup.select("div.base-search-card"): urn = card.get("data-entity-urn", "") # "urn:li:jobPosting:4321..." posted = card.select_one("time") jobs.append({ "job_id": urn.rsplit(":", 1)[-1], "title": text(card.select_one("h3.base-search-card__title")), "company": text(card.select_one("h4.base-search-card__subtitle")), "location": text(card.select_one(".job-search-card__location")), "posted_date": posted.get("datetime") if posted else None, }) return jobs def job_details(job_id): r = requests.get(f"https://www.linkedin.com/jobs/view/{job_id}/", headers=HEADERS, timeout=30) r.raise_for_status() soup = BeautifulSoup(r.text, "html.parser") criteria = {text(li.select_one("h3")): text(li.select_one("span")) for li in soup.select("li.description__job-criteria-item")} return { "seniority": criteria.get("Seniority level"), "employment_type": criteria.get("Employment type"), "industries": criteria.get("Industries"), "salary": text(soup.select_one("div.compensation__salary")), "applicants": text(soup.select_one(".num-applicants__caption")), "description": text(soup.select_one("div.show-more-less-html__markup")), } if __name__ == "__main__": jobs = search_jobs("python developer", "United States") print(f"{len(jobs)} jobs on the first page") for job in jobs[:2]: time.sleep(2) job.update(job_details(job["job_id"])) print(job["title"], "|", job["company"], "|", job["seniority"], "|", job["salary"], "|", job["applicants"], "|", len(job["description"] or ""), "chars of description") 

Running it today printed:

10 jobs on the first page Senior Python Software Developer (Onsite) | Raytheon | Mid-Senior level | None | Be among the first 25 applicants | 8807 chars of description Software Engineer II - Python & Cloud Services (Onsite) | Raytheon | Mid-Senior level | None | Be among the first 25 applicants | 8515 chars of description 

That's the whole idea: page through the search with start, collect the IDs, and open each job page for the details. For a few dozen jobs from your own machine, this is enough.

One gotcha: there is also a guest fragment for a single job, /jobs-guest/jobs/api/jobPosting/<id>, which looks like a cleaner source for the details. When I compared it with the real job page on 20 jobs, it said "Not Applicable" for seniority on 5 of them, where the page showed "Entry level" or "Mid-Senior level". The other fields matched. Use the job page.

Where the simple version breaks

These are the problems I hit when running this at volume from cloud servers. The numbers come from my own runs.

  1. Rate limiting. LinkedIn answers HTTP 429 to part of the requests that come from datacenter IPs: 13 to 15 % of them in my tests, sometimes from the very first request. You need a pause between requests, retries and a fallback (another IP or a proxy). Going faster doesn't help. One request at a time with a short pause gives me about 16 jobs per minute with full details.
  2. About 1,000 results per search. LinkedIn stops paginating around there. Split a big search by city or by date posted.
  3. The search is fuzzy. Since August 2026, a search for "python developer" also returns related titles. In today's run of 100 jobs, 49 had "python" in the title. Filter the titles yourself.
  4. Salaries are often only in the text. 23 of today's 100 jobs had LinkedIn's salary box, and 39 more gave the pay only in the description ("The base pay range for this role is..."). If salary matters to you, you need a parser for free text that also gets the currency and the period.
  5. Duplicates and "new since yesterday". The same job shows up in several searches, and for a daily alert you have to remember which jobs you already have.
  6. The markup changes. Class names like base-search-card__title have been stable for a while, but when they change, a scraper returns empty fields without failing. Check the counts, not just the exit code.

The hosted option

If you'd rather not maintain the retries, proxies, salary parsing and dedup, I run a hosted version: LinkedIn Jobs Scraper on Apify. Disclosure: I built it, and I get paid when people use it.

You call it through the Apify API. With the official Python client (pip install apify-client) and your API token in APIFY_TOKEN:

import json import os from decimal import Decimal from apify_client import ApifyClient client = ApifyClient(os.environ["APIFY_TOKEN"]) run_input = { "keywords": ["python developer"], "locations": ["United States"], "datePosted": "pastWeek", "maxJobs": 100, } # max_total_charge_usd caps what this run can cost, whatever the input. run = client.actor("tablewise/linkedin-jobs-scraper").call( run_input=run_input, max_total_charge_usd=Decimal("1.00") ) items = client.dataset(run.default_dataset_id).list_items().items for item in items: print(item.get("title"), "|", item.get("company"), "|", item.get("salary") or "-") with open("jobs.json", "w", encoding="utf-8") as f: json.dump(items, f, ensure_ascii=False, indent=2) 

Today's run (WasienIIfhwpzATNu) returned 100 jobs in a little over 6 minutes. One of the rows, with the description shortened:

{ "platform": "linkedin", "job_id": "4490678674", "title": "Full Stack Python Developer", "company": "A. Blair Enterprises, Inc.", "company_url": "https://www.linkedin.com/company/a-blair-enterprises-inc", "location": "Louisville, KY", "workplace_type": null, "workplace_type_source": null, "remote": null, "employment_type": "Full-time", "seniority": "Mid-Senior level", "industries": "Advertising Services", "salary": "$80,000 - $125,000", "salary_min": 80000, "salary_max": 125000, "salary_currency": "USD", "salary_period": "year", "salary_source": "description", "posted_date": "2026-11-16", "posted_text": "4 days ago", "applicants_count": 103, "applicants_text": "103 applicants", "easy_apply": false, "apply_on_company_site": true, "benefits": [], "url": "https://www.linkedin.com/jobs/view/4490678674/", "apply_url": "https://www.linkedin.com/jobs/view/4490678674/", "description": "Full Stack Python Developer (AI-Accelerated)\n\nCompany: A. Blair Enterprises\n\nLocation: Louisville, KY 40223 (In-person)\n\nCompensation: $80,000 - $125,000 a year ...", "search": "python developer in United States", "search_position": 55, "status": "ok", "status_message": null, "scraped_at": "2026-11-21T14:18:17Z" } 

What it adds over the script above:

  • salary from the pay box or from the description, as salary_min, salary_max, currency and period, with salary_source saying where it came from;
  • workplace_type, read only from clear phrases ("fully remote", "Work Arrangement: Hybrid"), with its source;
  • whether you apply with Easy Apply or on the company site;
  • titleMustInclude and titleExclude, so the fuzzy matches are dropped (and not charged);
  • onlyNewJobs for scheduled runs, so a daily alert only returns, and charges for, jobs you haven't seen yet.

Cost: $1 per 1,000 jobs plus $0.0005 per run, with the details included. At that price, today's 100 jobs come to $0.1005. A search that finds nothing adds one free status row. Apify's free plan includes $5 of credit every month, which is about 4,990 jobs.

The same call without a client library:

curl -X POST "https://api.apify.com/v2/acts/tablewise~linkedin-jobs-scraper/run-sync-get-dataset-items" \ -H "Authorization: Bearer $APIFY_TOKEN" \ -H "Content-Type: application/json" \ -d '{"keywords": ["python developer"], "locations": ["United States"], "maxJobsPerSearch": 20}' 

That endpoint waits up to 5 minutes. For bigger runs, start the run and read its dataset when it finishes, as the Python example does.

Is this legal?

The data here is public: job postings that anyone can see without logging in. Still:

  • LinkedIn's User Agreement, which members accept, prohibits scraping, and its robots.txt disallows every crawler except approved search engines.
  • In hiQ v. LinkedIn, US courts said that scraping public pages is likely not "unauthorized access" under the CFAA. hiQ, which had LinkedIn accounts, was later found to have breached the User Agreement.
  • Job descriptions sometimes contain personal data, such as a recruiter's name or email. If you store it, data protection laws like the GDPR apply.

It depends on where you are and what you do with the data. If it matters for your project, ask a lawyer.

Wrap-up

  • For a quick look at a few dozen jobs, the short script above is enough.
  • At volume, the hard parts are rate limits, fuzzy search, salary parsing and dedup, not the HTML.
  • The Python and JavaScript versions of the hosted example are in this examples repository.

This article was written with AI assistance. All the code in it was run before publishing, and every number comes from real runs on 21 November 2026 (run ID above) or from earlier test runs of the hosted scraper.