Indeed Jobs Scraper
Pricing
from $0.02 / 1,000 jobs
Scrape Indeed jobs in 60+ countries by keyword, location or URL, beyond the 1,000-job limit: title, full description, salary, location, job type, remote, company profile and rating. Export to JSON, CSV or Excel.
Indeed URLs from any country site: search results (https://www.indeed.com/jobs?q=nurse&l=Texas, keyword, location, radius, date, sort and job-type filters are kept), company job pages (https://www.indeed.com/cmp/Hubspot/jobs; its own keyword and place are kept: https://www.indeed.com/cmp/Hubspot/jobs?q=engineer&l=Boston) or single jobs (https://www.indeed.com/viewjob?jk=4f9a1c2b3d4e5f60). When this list is not empty, the search fields below (keywords, locations, countries, companies) are ignored; filters, caps and monitoring still apply. For a single job link, the date, job-type, work-arrangement, experience, pay and language filters are checked on the job itself (a job that does not show the criterion is dropped, and counted as filtered out); the radius and the location of a search cannot be read on a job and are not applied. Max 1 000 URLs.
[]Job title, skill or company, as typed on Indeed (e.g. software engineer). Indeed search syntax works: title:(nurse), "exact phrase", -excluded, or. Empty = every job of the location.
Several keywords in one run: one search per keyword (times each location and each country). Added to Search keyword; jobs found by several searches are saved once.
[]City, state, region or postal code, as typed on Indeed (e.g. New York, NY, Paris, Remote). Empty = the whole country.
Several locations in one run: every keyword is searched in every location (3 keywords × 4 locations = 12 searches per country, max 500 searches). Added to Location.
[]Indeed country site to search (each country has its own jobs). With no keyword, no location, no URL, no job key and no company: the default (United States) alone runs a small example (20 jobs), any other country reads every job of that country (up to Max jobs). The labels (tags, sectors, pay text) come in the language of the country's site, as on Indeed.
Search several Indeed country sites in one run: every keyword × location is searched in each of them. With Country left on United States, only the countries of this list are searched (to include the US, put US here); another Country is added to the list. Leave Location empty to cover each whole country. With no keyword or location, a filled list alone reads every job of those countries (the default country is then not added: for the whole US, put US here).
[]All the open jobs of these employers: company names (Starbucks), website domains (starbucks.com), Indeed company pages (https://www.indeed.com/cmp/Starbucks) or employer keys. A name or domain is matched to the exact Indeed employer (a domain must match the employer's website), not to look-alike names. Combined with the keywords and locations above when they are filled.
[]Indeed job keys (the jk= value of a job link) to read directly, up to 10 000. They are read on top of the keyword and location searches: empty those fields to read only these jobs. Expired jobs are returned too, flagged as expired. The date, job-type, work-arrangement, experience, pay and language filters are checked on each job (a job that does not show the criterion is dropped, and counted as filtered out); the radius and the location cannot be read on a job and are not applied. A job link is read on the site of the link, a bare key on the Country site — on the first of Countries when Country is left on United States and that list is filled (give the country of the job to get its pay).
[]Maximum number of jobs to save for the whole run (after deduplication and filters). 0 = no limit.
Cap for EACH search (keyword × location × country, company, or search URL), so that the first search cannot use up the whole Max jobs budget. 0 = no per-search cap.
Indeed shows at most 1,000 jobs per search. When a search has more, the Actor splits it into publication-date slices and collects each one (measured: 99.9 % of searches of 2,500 to 13,700 jobs instead of the first 1,000). The log and the COVERAGE record give announced vs collected jobs per search. Off = first 1,000 only.
Which forms of the full job description to save. Smaller = lighter dataset.
Add the e-mail addresses and phone numbers the employer wrote in the job description (emails, phones). Charged only for jobs where at least one is found (on 700 jobs read in the US, France and Germany: 19 % had an e-mail, 8 % a phone number). See Pricing.
Also save one row per employer found (recordType = company): profile, total open jobs on Indeed, new jobs in the last 24 hours / 7 days / 30 days, and the change since the previous run with the same Memory key. Ready for a CRM or n8n. Charged per company row. See Pricing.
Only jobs added to Indeed in this period.
Keep jobs of ANY of these types. Empty = all. Outside the USA, the fixed-term contract Indeed shows as "Contract" (CDD, befristet…) is Fixed-term contract; Contract is the contractor, mostly USA, Canada and Australia.
[]Work arrangement as tagged by Indeed.
Keep jobs of ANY of these levels, as tagged by Indeed. Empty = all.
[]Keep jobs of employers in ANY of these sectors, as named by Indeed in the country's language (e.g. Healthcare, Information Technology, Banking) — matched on the employer's sector — or by Indeed sector key (e.g. 22VKN), which Indeed itself filters on. Use either names or keys in one run, not both. Empty = all.
[]Jobs whose yearly pay range reaches this amount, in the country's currency (Indeed's own filter: includes salaries estimated by Indeed). 0 = no filter.
Jobs whose yearly pay range starts below this amount, in the country's currency (Indeed's own filter, like Minimum yearly salary). 0 = no limit.
Distance around Location. 0 = exact place only. Empty = Indeed's default (25 miles).
Only jobs written in this language: two-letter code such as en, fr, de. Empty = all.
Order of the results (matters with Max jobs). With Only new jobs or Posted after, the results are always read newest first.
Only jobs added to Indeed on or after this date (dateOnIndeed): 2026-09-01 (a date alone starts at 00:00 UTC), or a period before now such as 7 days, 2 weeks, 1 month (API: 24 hours and full ISO date-times work too). Also sent to Indeed to read fewer pages.
Only jobs added to Indeed on or before this date (the whole day is included, in UTC), or older than a period such as 30 days.
Drop the jobs whose title, or the cleaned-up title Indeed shows (displayTitle), contains one of these words (case and accents ignored).
[]Drop the jobs of these employers (the company name as shown, its Indeed company page name companyProfileName, or its parent group parentCompanyName, contains one of these words, case and accents ignored), e.g. staffing agencies.
[]Drop the jobs that show no pay.
Indeed Apply = jobs candidates apply to on Indeed (the isEasyApply column); Employer site = jobs that send candidates to the employer's or ATS page (the applyUrl column).
Drop the jobs of employers rated below this on Indeed (out of 5). Jobs of unrated employers are dropped too when set. 0 = no filter.
Skip the jobs that a previous run (same Memory key) already delivered: they are not saved and not charged. First run = everything is new.
Name of the memory used by Only new jobs and by the company change counts. Give each schedule / task its own key (e.g. nurses-texas) so that they do not share their memory. Letters, digits, - and _.
Forget everything remembered under this Memory key before the run: this run returns (and charges) every job again. Untick it afterwards.
Apify Proxy or your own proxies. Keep the default: it is included in the price. The residential Apify proxy is not available in this Actor.
{ "useApifyProxy": true}Maximum number of requests processed in parallel. Lowered by itself to what the memory of the run holds (5 at the default 512 MB, 17 at 1 GB): give the Actor more memory to read more at a time.
Most requests to Indeed in any 60 seconds (each request returns up to 100 jobs). A budget, not an even pace: up to this many can leave at once when the minute starts (spread them with the minimum delay below). Lower it if the log shows HTTP 429 / 403.
Smallest gap between two requests, in milliseconds. Unlike the per-minute rate, this spreads the requests evenly instead of letting them go out in a burst. 0 = no gap.
Retries per request before it is marked as failed. Behind a proxy, a request the site turns away is also retried on a new proxy session up to 10 times without using up these retries.