LinkedIn Jobs Search Scraper API - Filters That Hold

job-detail

One LinkedIn job you named by its URL or id, delivered with its full posting: title, company, location, posted time, the complete description as clean text and as HTML, seniority level, employment type, job function, industries, applicant count where LinkedIn prints one, and salary when the posting states one. One request per row, which is why it costs more than a job found by a search. A job that has closed or was removed costs nothing and returns a free row saying which.

job-card

The row you get when you turn off 'Read the full posting': id, URL, title, company, location, exact posted date, badges such as Actively Hiring, and the logo, taken from the search listing without opening the posting. One request covers ten of these, which is why it costs a third of a full job row.

company

One LinkedIn company or showcase page delivered: name, tagline, description, website, industry, size band, the exact number of members who list the company, exact follower count, headquarters and every office address, organisation type, founding year, specialties, logo and cover, the numeric organisation id, affiliated and similar pages, the employees LinkedIn features, and the ids of the ten newest posts on the page. On a job run with 'Add company pages' on, charged once per distinct company in the run, never per job. A company page that does not exist costs nothing.

post

One public LinkedIn post delivered, whether read from its own page or from the newest posts on a company page: the full text, exact reaction and comment counts, the author with their exact follower count, when it was published, media type, image URLs, the native video URL, the linked article, the original post when it is a reshare, hashtags and mentions. A post that has been deleted or that LinkedIn does not show publicly costs nothing and returns a free row saying which.

comment

Optional. With 'Include the comments shown on the post' on, each comment LinkedIn renders on the public post page is its own row: the commenter's name and profile URL, the text, when it was posted and its like count. LinkedIn shows about ten comments publicly; the post row always carries the exact total. Charged only for comments actually delivered.

article

One LinkedIn Pulse article delivered: title, summary, the full body as clean text and as HTML, cover image, published and modified dates, reading time, exact like and comment counts, the author with their exact follower count, the newsletter it belongs to, and the comments shown on the page. An article that no longer exists costs nothing.

profile

One public LinkedIn profile delivered from its public page: name, headline, location, about, current positions, experience with company links and dates, education, awards, volunteering, publications, courses, websites, the exact follower count, connections, badges, the person's recent posts and articles, and which sections LinkedIn exposed. LinkedIn shows about 7 in 10 ordinary profiles to visitors without an account (measured 2026-09-02); a profile it declines costs nothing and returns a free row saying so. Charged only for profiles actually delivered.

search

One keyword-and-location job search actually run against LinkedIn and read. Charged once per search that returned listings, never for a search that came back empty, and never more than 250 times in one run however many searches you cross. It never counts against your result cap. This is what keeps a scheduled search worth running on a quiet day: with change-only billing on, a day with no new postings costs you this and nothing else.