Company career page jobs: the job openings of the companies you name, live from their Lever, Greenhouse, Ashby, Workable, Workday, Recruitee, Personio or Teamtailor boards: title, location, pay where published, link. Input: board links or names. $1.00 per 1,000 jobs on Free, $0.70 on Business.
README: a short note and Apify's video on running it on a schedule.
README: related actors updated (same-topic links first, current names).
0.1.34 - 2026-11-15
A run that fails as a whole now writes a full summary built from what it did up to the failure (companies, jobs saved, unfinished companies, overviews), marked failed with the reason and with alerts marked not sent, instead of a bare error record.
README links a ready-made Google Jobs search example on GitHub.
Added requisitionId where available and per-company jobsOverview in SUMMARY with saved postings, postings with a requisition ID, distinct requisition IDs and distinct titles, top departments and locations, and newest dated postings.
0.1.33 - 2026-11-15
FAQ: "Is there a Google Jobs API?" Google's Cloud Talent Solution searches jobs you upload yourself; this Actor works as a job scraper API (Google Jobs search on paid plans, US).
README opens with a first check: the daily only-new watch task, what it costs (up to 300 jobs: at most $0.30 on Free, $0.21 on Business, plus $0.00005 per run) and one real saved job.
README now names the Actor "Company Career Page Jobs Scraper", matching its Store title (job openings from company career pages and, on paid plans, US Google Jobs).
Added optional Google Jobs keyword search at the existing job price: paid plans only, US only for now (non-US not measured). Free-plan and local runs show an availability notice and continue with requested career-site work.
Rows identify their source and Google Jobs can supply application choices, a listing link, posting age, employment types, published labels, salary text and a description when requested. Job highlights, related links and a company logo are not provided by this option.
Google Jobs rows that may duplicate a career-site posting are kept, flagged in possibleDuplicateOf and charged as separate rows. The flag requires a retained application link matching a career job/apply URL plus equal company names after case and whitespace folding; URL comparison ignores tracking parameters, fragments and trailing slashes. Filter possibleDuplicateOf is null to exclude flagged rows from exports (saved rows remain charged), or do not select google_jobs together with career-site systems or list career-site boards in that Google Jobs run. The summary reports flaggedPossibleDuplicates.
Repeated Google job IDs are deduplicated, up to 20 distinct application-link aliases per ID survive restarts in first-seen order (including application choices); later aliases are omitted, and watch memory remembers returned Google IDs across runs. Career-site processing reads boards as needed and preserves its existing output when Google Jobs is not selected. Career-site per-company caps use canonical career boards, including separate allowances for boards sharing a company name and one allowance for different hiring entities on one board. Google Jobs uses separate per-company caps keyed by the company name Google writes, with case and whitespace folded; Google rows do not consume or block career-board allowances. Alias processing checks the remaining run time between representations. Shared job and charge caps, paid-plan access checks, refusal handling and the three-consecutive-failure stop remain in place.
0.1.32 - 2026-11-14
The Actor is now named Company Job Openings Scraper (its address, lever-jobs-scraper, is unchanged). The listing and README opening state the Free-plan price and a paid-plan price and the input it needs, and name the eight job-board systems (Lever, Greenhouse, Ashby, Workable, Workday, Recruitee, Personio, Teamtailor). The README also names the Workday lookup by company name, Workday and Recruitee pay, and how a watch treats an incomplete answer. No change in behaviour or price.
0.1.31 - 2026-11-13
README: a "Use it in a pipeline" section with one step-by-step recipe for Clay (hiring signals in a table), using Apify's own integration and your own Apify account. It says what it was checked against and that it was not run in a live account. The how-to steps now say that a company name alone is looked up on the boards and on Workday. New FAQ answer on downloading the results as CSV, Excel or Google Sheets. No change to the Actor, its input or its output.
0.1.30 - 2026-11-13
Maintenance: the Personio hiring entity's trailing number is removed in one pass, so a feed with a very long entity name can no longer slow a run down. Results are unchanged.
0.1.29 - 2026-11-13
Maintenance: descriptions (as text, HTML and Markdown), pay ranges, block-page checks and the keyword search's result addresses are read in one linear pass. A description or page with many unclosed tags, or with one very long word, can no longer slow a run down. Results are unchanged.
0.1.28 - 2026-11-13
Maintenance: the Workday addresses shown in search results are read in one linear pass, so a page with many unclosed tags can no longer slow a run down. Repeated Workday addresses are now merged in one pass too, so a result page naming tens of thousands of distinct sites stays fast. Results are unchanged.
0.1.27 - 2026-11-13
Maintenance: the check that stops web searches on a verification or consent page now also reads the page's tags the way a browser does, so a check form or widget written with spaces around "=", without quotes, or after a quoted ">" is recognised too; everything the earlier check caught is still caught. The rest of that check (removing tags, reading headings, counting result links) runs in one linear pass. Results on ordinary pages are unchanged.
0.1.26 - 2026-11-13
Maintenance: code blocks in descriptions and the check of web-search answers for verification pages are read in one linear pass, so a page with many unclosed tags can no longer slow a run down. Results are unchanged.
0.1.25 - 2026-11-12
Run summary: keywordSearch now counts the boards actually read (boardsRead), so a run that stops at your maximum jobs no longer reports every chosen board as read. New per keyword: boardsSelected (chosen, at most 10) and boardsLeftUnread (chosen but not read because the run stopped first); boardsRead for the whole run is added. boardsFound is unchanged. A board closed by a service's earlier refusal, without a request of its own, counts as unread. After a platform restart boardsRead counts only the boards requested after the restart, and boardsLeftUnread is the boards selected minus that count.
0.1.24 - 2026-11-12
Documentation: the README, input help and output description were updated to cover all eight systems, the keyword search, the seniority and workplace filters, the seniority and no_seniority filter counts, and the company reason codes access_state_unsaved, aborted, lost_watch_lock and rejected_request; Recruitee also states an update date for Updated since.
0.1.23 - 2026-11-12
If the record of a refusal or a pause cannot be saved with the run, the run now ends at once and says so in its log, status and run summary (stoppedByaccess_state_unsaved, accessStateNotSaved). Before, it stopped asking but went through its remaining companies, reporting each as failed.
0.1.22 - 2026-11-12
Recruitee boards are read from the company's job feed instead of the careers-site API, which needs a company token from 10 February 2027. The rows are the same: on three boards read on 12 November 2026, every one of 379 jobs gave the same row from both, field for field, pay, locations, workplace, dates and descriptions included.
0.1.21 - 2026-11-12
Every row has seniority, read from the job title only by fixed rules: intern, junior, mid, senior, lead, principal, director or manager; null when the title states no level. It is never taken from the description and never guessed. The rules, with the words that only look like a level, are in the README.
New filter: keep only jobs of the seniorities you choose. Jobs whose title states no level are left out unless you keep them; jobs left out are counted in the run summary and not charged. Other actors' seniority and experience_levels inputs are read as this filter.
0.1.20 - 2026-11-12
Find job boards by keyword: give keywords (and optionally a location and the systems to search) instead of, or as well as, company names. Boards of companies you do not name are found by a web search and read live; only their jobs that match the keyword are saved, and each row says which keyword in foundByKeyword. Finding boards is free; jobs are charged as before. The boards are those the search engine has indexed, about ten per keyword and system, so some can be missed.
At most 20 web searches per run, shared with the lookup of company names on Workday, and at most 10 boards read per keyword (the rest are counted in the run summary). A web search is never sent twice, also when the platform restarts a run.
Other actors' "search all companies" input with keywords is now read as this keyword search.
0.1.19 - 2026-11-12
Input help: a company name alone is also looked for on Workday (as since 0.1.13).
0.1.18 - 2026-11-12
Feeds nested unusually deeply are rejected as invalid instead of being read.
0.1.17 - 2026-11-12
Three more job-board systems: Recruitee, Personio and Teamtailor. Give a company's careers link (<company>.recruitee.com, <company>.jobs.personio.de or .com, <company>.teamtailor.com) or recruitee:<company>, personio:<company>, teamtailor:<company>, and its open jobs come back in the same rows as the other systems. A bare company name is still looked up on Greenhouse, Lever, Ashby, Workable and Workday only.
Recruitee jobs carry the pay range the employer publishes (amounts, currency and whether per hour, month or year), so the salary filters work on them; Personio and Teamtailor publish no pay. Personio jobs give the hiring legal entity as the company name and the schedule (full-time, part-time) as the employment type.
A Teamtailor careers site is read only where its robots.txt allows it; a site that does not allow it, or whose robots.txt cannot be read, is skipped and costs nothing. The run summary states each Teamtailor site's stated content uses (for example ai-train=no).
As for every system, no recruiter, hiring-manager or contact field is returned, and e-mail addresses and phone numbers are removed from descriptions.
In HTML and Markdown descriptions, on every system, a link whose address holds an e-mail address or a phone number is given as its text only, without the address.
0.1.16 - 2026-11-12
Links in Markdown descriptions are always written as one plain link.
0.1.15 - 2026-11-12
Keywords can now also be matched in job descriptions (Match keywords in: title and description), without turning descriptions on and at no extra cost; on Workday boards keywords are still matched in titles only.
New workplace filter: remote, hybrid or on-site. Jobs whose workplace is not stated are left out and counted, or kept if you choose.
Descriptions can be returned as plain text, HTML or Markdown, at no extra cost.
On an only-new watch, a job the watch returned that has left its board is now listed once as closed, free, in the run's CLOSED_JOBS record; this never happens for a board that was not read completely.
Field names other job-board actors use for these options are read as this Actor's own.
0.1.14 - 2026-11-12
A company found on Workday by name is now read more often: when only its Workday address is known, its main external careers site is read if the company's own site makes it clear which one that is.
0.1.13 - 2026-11-12
A company name alone now also finds the company's Workday careers site when it has no board on Lever, Greenhouse, Ashby or Workable (up to 20 names per run, nothing charged for the lookup), so "Boeing" or "General Motors" works without the careers link.
0.1.12 - 2026-11-12
Very long job descriptions are read for pay ranges in bounded time, and a second range stated anywhere in a description, also without a pay word or currency beside it, leaves the pay amounts empty.
0.1.11 - 2026-11-12
Workday jobs read with descriptions now carry the pay range the employer writes in the description (amounts, currency and whether per year or per hour); a description stating several different ranges lists them without picking one.
0.1.10 - 2026-11-12
A Workday site read only up to your maximum is now reported as capped in the run summary, not complete, with the number of jobs the site lists beside it.
A watch on a Workday site treats a list as complete only when every job the site states was read, so a job it has not seen is never taken as closed and returned again later.
0.1.9 - 2026-11-12
Workday careers sites: paste a company's careers link (https://.wd.myworkdayjobs.com/, with or without a language) or workday:.wd/, and its open jobs come back in the same rows as the other systems. Each site is read only where its robots.txt allows it; a site that does not allow it, or whose robots.txt cannot be read, is skipped and costs nothing. Bare company names are not looked up on Workday.
Workday states when a job was posted only relatively ("Posted 2 Days Ago"): each row keeps those words (postedText) and the date counted back from the day of reading; "Posted 30+ Days Ago" gives no date. With descriptions on, each Workday job's details are read (one request per saved job), which also give every place, the country, the time type and the exact posting day.
A Workday site shows at most 2,000 jobs, newest first; the run summary states each site's own count and whether its list was read to the end, and a watch never forgets a job that drops out of the 2,000 shown. With descriptions on, filters on place, type, pay and dates are checked on what each job's details state, so nothing that does not match is saved or charged.
0.1.8 - 2026-11-12
Workable job boards: give a company's careers link (apply.workable.com/) or workable:, and its open jobs come back in the same rows as the other three systems, with the company name Workable states. Workable publishes no pay range and no update time, so those fields stay empty for its jobs; remote comes from Workable's own remote flag. A bare company name is now looked up on all four systems.
Every row now lists the title keywords it matched (matchedKeywords), so you can check your keyword filter on each job; the run summary already counts, per company, the jobs on the board, those that matched, and those each filter left out.
0.1.7 - 2026-11-12
Only-new watch: when a run reaches your maximum jobs per run, the companies it had not reached yet are still read once, so their open jobs are remembered without being saved or charged. A later run (with one run of the watch at a time) then returns only jobs that appeared since, at every company. A company the watch has not read yet, because the maximum charge, the timeout or a failure stopped a run before it, gets its first read, up to your maximum, in the run that first reads it.
New filters: employment types as the board states them, and a minimum or maximum salary and a salary currency, compared with the published pay range in the job's own currency and interval. Jobs left out by them are not charged; jobs without a published pay range are left out while a salary filter is set.
The listing links four ready-made examples and states its measured speed, memory and maximum-charge figures from one set of runs.
0.1.0 - 2026-11-12
First public version: every open job of the companies you name, read live from their own Greenhouse, Lever and Ashby job boards. Give a job board link or a name with its system, such as greenhouse:airbnb. A bare name is looked up on all three systems and the board with the most open jobs is read; the others are listed in the run summary.
One row per job: company and, where the board states it, its display name, title, department and team, locations, country, workplace type and remote flag as the employer states them, employment type, pay range with currency and interval when published, posted and updated dates, job and apply links, and optionally the full description as text with e-mail addresses and phone numbers removed.
Filters on title keywords, excluded words (whole words), locations (one place per line), departments, remote jobs only and posting or update dates. Jobs left out by a filter are not charged.
A watch for scheduled runs (one run at a time): the first run returns up to your maximum of the newest matching jobs and remembers every job open on each board; later runs return only jobs that appeared after. Older matches are not back-filled. Alerts to Slack, Discord or a webhook. Companies can also be read from another dataset.
$1.00 per 1,000 jobs on the Free plan ($0.70 on Business and above), plus Apify's start event. Maximums per run and per company, and the run's maximum charge, are respected; a company without a board, a board that cannot be read and jobs a run did not reach are not charged, and a job is never saved twice, also when saving it reports an error. The run summary says per company what was read, saved, left out and why.
Limits: only companies whose careers page uses Greenhouse, Lever or Ashby; boards on the EU instance of Greenhouse or Lever are not read yet; at most 100 bare names per run. Recruiter and hiring-manager fields are never returned.