Google Jobs Scraper API - New Jobs, Pay per Job

Google Jobs Scraper extracts job listings from Google Jobs for any job title, location, country and language: title, company, location, source site, full description, posted date, salary, employment type and every apply link. Use it as a Google Jobs API for job market research, recruiting leads or a job board feed. Monitoring mode returns only new jobs since your last run, so scheduled runs cost only what changed. Filters for date posted and employment type are applied by Google and checked again on every job, and staffing agencies can be excluded. You pay only per unique job saved: empty searches, duplicates, excluded jobs and failed searches are free, and there is never a charge per Google page.

Why this Google Jobs scraper

• Pay per unique job, never per page. One flat price per job saved. No charge for empty results, duplicates, errors or the extra pages needed to apply filters or go deeper.
• Only new jobs since the last run. For daily or hourly schedules, the Actor remembers the jobs delivered for each query and saves only new ones.
• Filters that are really applied. Date posted (since yesterday, last 3 days, last week, last month) and employment type go through Google's own filters, then are checked again on every job before it is saved.
• Exclude staffing agencies and companies, or keep only the ones you want. Drop agency postings with one switch, exclude any company by name, or keep only target companies, jobs published on a given site (LinkedIn, Indeed, the company career site) or jobs that show a salary.
• More than 10 results per query. Google Jobs shows 10 jobs per page. The Actor follows Google's refinement links (date windows, job types, Remote and similar) and removes duplicates, so one query returns several times more unique jobs.
• Clean, normalized data. ISO 8601 dates, salaries split into minimum, maximum, currency and period, standard employment type codes and all apply options with their real URLs.
• Any country, eight languages. 43 country editions and 8 interface languages (English, Spanish, Portuguese, French, German, Italian, Dutch and Polish), with dates and job types normalized in each.
• Reliable by design. No browser. Every page is read through the Apify Google SERP proxy, with retries, exponential backoff and a retry budget per run. Progress is saved, so a platform restart never duplicates rows or charges twice.

Looking for an epctex Google Jobs Scraper alternative or a replacement for a Google Jobs Actor that stopped returning data? This Actor takes a query, location, country and date filter, reports a clear status for every query and charges only for jobs actually delivered.

How to use it

1. Enter one or more job titles or keywords, one per line (for example data analyst, registered nurse).
2. Enter a location such as Austin, TX, London or Bogotá, or leave it empty to search the whole country.
3. Choose the country and language. Use the language of your queries (Colombia and Spanish for desarrollador).
4. Optional: set Date posted, Employment type, Exclude companies, Exclude staffing agencies, Only these companies, Only jobs published on, Only jobs with salary and Max jobs per query. Jobs removed by a filter are not saved and not charged.
5. For recurring runs, enable Only new jobs since last run.
6. Run the Actor and download the results as JSON, CSV or Excel, or read them through the API.

Example input for a daily monitor:

{
  "queries": ["data analyst", "business intelligence analyst"],
  "location": "Austin, TX",
  "country": "US",
  "language": "en",
  "datePosted": "3days",
  "excludeStaffingAgencies": true,
  "onlyNewSinceLastRun": true,
  "monitoringStoreName": "austin-analysts",
  "maxResultsPerQuery": 100
}

Monitoring: only new jobs since the last run

With Only new jobs since last run, the Actor keeps a named key-value store (default name google-jobs-monitor) that is kept between runs. For each query, combined with its location, country, language, date filter and employment type, it remembers the Google job IDs already delivered. Later runs skip those jobs, so they are neither saved nor charged.

• The first monitored run of a query returns all matching jobs. Its RUN_SUMMARY entry shows firstMonitoredRun: true.
• When nothing is new, the query ends with status NO_NEW_JOBS, which is not a failure, and the run costs nothing.
• Changing the location or a filter starts a separate memory for that combination, so different monitors never interfere.
• Use a different Monitoring store name for each independent monitor, for example one per client. Delete the store in Storage to start over.
• Recommended: combine monitoring with Date posted set to "Last 3 days" or "Last week". New listings appear first in Google's recent-jobs pages, and the Actor stops early when pages bring nothing new, which keeps runs short.

Integrations: schedules, API, webhooks, Sheets, Zapier, Make and AI agents

Schedules. Save your input as a task, then open Schedules in Apify Console and run it daily or hourly. With monitoring enabled, each run returns only new jobs.

API. Run the Actor and get the jobs in one call:

curl -X POST "https://api.apify.com/v2/acts/ocius-data~google-jobs-reliable/run-sync-get-dataset-items?token=YOUR_API_TOKEN&format=json" \
  -H "Content-Type: application/json" \
  -d '{"queries": ["registered nurse"], "location": "Houston, TX", "datePosted": "today"}'

Python, with the official apify-client package:

from apify_client import ApifyClient
client = ApifyClient("YOUR_API_TOKEN")
run = client.actor("ocius-data/google-jobs-reliable").call(run_input={"queries": ["python developer"], "location": "London", "country": "GB", "onlyNewSinceLastRun": True})
for job in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(job["title"], job["company"], job["applyLink"])

The latest run's jobs are always at https://api.apify.com/v2/acts/ocius-data~google-jobs-reliable/runs/last/dataset/items?token=YOUR_API_TOKEN&status=SUCCEEDED (add &format=csv for CSV).

Webhooks. In the Actor or task, open Integrations, add a webhook for ACTOR.RUN.SUCCEEDED and point it at your endpoint. The payload includes resource.defaultDatasetId to download the new jobs.

Google Sheets. Add the Google Sheets integration in the Integrations tab of your task, or import the CSV export. Use jobId as the unique column.

Zapier, Make and n8n. Use the official Apify app or node. Trigger: "Actor run finished". Action: "Get dataset items". Then post each new job to Slack, email, Airtable or your CRM.

MCP and AI agents. The Actor runs with limited permissions, so it is available through the Apify MCP server (https://mcp.apify.com). Add ocius-data/google-jobs-reliable to the server's Actor list and an AI agent can call it with plain inputs such as {"queries": ["nurse"], "location": "Miami, FL"}.

Pricing and cost calculator

Event	Price	When it is charged
job	USD 0.003 per job	One unique job saved to the dataset, with full description, highlights and all apply links included.

There is no start fee and no charge per Google page; proxy costs are included. Never charged: searches with no results, duplicates, jobs removed by your filters or exclusions, jobs already delivered in monitoring mode, and failed searches. Your maximum cost per run is respected: the Actor stops as soon as it is reached and never saves a job it cannot charge. The Actor runs with 1 GB of memory by default.

Scenario	Unique jobs saved	Cost
Quick check: 5 queries, 20 jobs each	100	USD 0.30
Weekly market scan: 20 queries, 50 jobs each	1,000	USD 3.00
Large scan: 100 queries, 100 jobs each	10,000	USD 30.00
Daily monitor of 10 queries, about 15 new jobs per query per day	150 per day	USD 0.45 per day, about USD 13.50 per month
Same monitor on a day with no new jobs	0	USD 0.00

Monitoring mode. When Only new jobs since last run is enabled, each query check costs USD 0.01 (event monitor-check), whether or not new jobs are found, because every check queries Google. New jobs are charged as usual. Example: one query checked every day for a month costs about USD 0.30 plus the new jobs delivered.

Output example and field dictionary

{
  "query": "data analyst",
  "searchLocation": "New York, NY",
  "jobId": "KdTqnnhBcP9XKf20AAAAAA==",
  "title": "Data Analyst, Payments",
  "company": "Current",
  "location": "New York, NY",
  "via": "Indeed",
  "postedAt": "2026-10-29T19:00:00.000Z",
  "postedDate": "2026-10-29",
  "postedAtRaw": "3 days ago",
  "salaryRaw": "130K–175K a year",
  "salaryMin": 130000,
  "salaryMax": 175000,
  "salaryCurrency": "USD",
  "salaryPeriod": "YEAR",
  "employmentType": "Full-time",
  "employmentTypes": ["FULL_TIME"],
  "applyLink": "https://www.indeed.com/viewjob?jk=e8f137077a618a54",
  "applyLinks": [{ "title": "Apply on Indeed", "source": "indeed.com", "url": "https://www.indeed.com/viewjob?jk=e8f137077a618a54" }],
  "description": "Current is a leading US-based fintech company...",
  "googleJobsUrl": "https://www.google.com/search?q=data+analyst+New+York%2C+NY&udm=8&hl=en&gl=us",
  "scrapedAt": "2026-11-01T19:18:46.172Z"
}

Field	Meaning	When it is null
query, searchLocation	The query that found the job and the location you entered.	searchLocation when no location was entered
jobId	Google's job identifier. Stable across runs, the key for deduplication and monitoring.	Never
title, company	Job title and hiring company as shown by Google.	Never
location	Job location, for example Austin, TX or Anywhere for remote jobs.	When Google shows none
via	Site Google took the listing from, for example LinkedIn or Indeed.	When Google shows none
postedAt, postedDate	Estimated publication time (ISO 8601 UTC, one-hour precision) and its date part.	When Google shows no date
postedAtRaw, postedAtIsApproximate	Date as Google shows it (3 days ago, hace 2 días) and true for lower bounds such as 30+ days ago.	When there is no date
salaryRaw	Salary as shown, for example 115K–180K a year.	When the posting has no salary
salaryMin, salaryMax	Salary range as numbers. From 25 an hour gives only a minimum.	When there is no salary or bound
salaryCurrency, salaryCurrencyInferred	ISO 4217 code, and true when the currency was taken from the selected country.	When there is no salary
salaryPeriod	HOUR, DAY, WEEK, MONTH or YEAR.	When not stated
employmentType, employmentTypes	Type as shown and normalized codes (FULL_TIME, PART_TIME, CONTRACTOR, INTERN, TEMPORARY, PER_DIEM, VOLUNTEER).	Empty when not stated
tags	Other attributes Google shows, such as benefits.	Empty when none
applyLink, applyLinks	First apply option, and all options with title, source and url.	Empty when none
description, highlights	Full description and the Qualifications, Responsibilities and Benefits lists.	When disabled in the input
country, language, filters	Settings used for the search.	Never
googleJobsUrl, sourceUrl, scrapedAt	The same search in Google Jobs, the exact page read, and the collection time (UTC).	Never

The output tab has three views: Jobs, Salary and Details.

Statuses and errors

Each query gets one status in the RUN_SUMMARY record, with requested, delivered, pagesFetched, stopReason, seenInPreviousRuns (monitoring), excludedByCompany and a message. Status objects are never written to the dataset.

Status	Meaning	Charged	What to do
OK	Jobs were saved. If fewer than requested, stopReason explains why.	Only saved jobs	Nothing
NO_NEW_JOBS	Monitoring: every job found was delivered in an earlier run.	No	Nothing
NO_RESULTS	Google Jobs has no listings for this search (confirmed with a second request).	No	Broaden the query, location or date range.
BLOCKED	Google kept refusing after all retries, or the Google SERP proxy is not available.	No	Run again later.
SOURCE_ERROR	Google answered, but a filter was not offered or the page could not be read.	No	Check the language and location.
INVALID_INPUT	The query is not valid, for example longer than 200 characters.	No	Fix the query.
NOT_PROCESSED	Your maximum cost per run or the timeout was reached first. Not a failure.	No	Raise the limit or split the queries.

Stop reasons: NO_MORE_RESULTS, DIMINISHING_RETURNS (three pages in a row with fewer than two new jobs), PAGE_LIMIT, MAX_RESULTS_REACHED, CHARGE_LIMIT_REACHED and TIMEOUT. If more than 20% of queries fail, the run is marked as failed. Invalid settings stop the run within seconds with a message that says how to fix them.

Known limits

• 10 jobs per Google page. The Actor goes beyond 10 by following Google's refinement links. Measured on 1 November 2026: software engineer in Austin, TX returned 121 unique jobs with 25 pages; with the default of 15 pages a busy search returns roughly 80 to 100. Smaller markets and narrow filters return fewer.
• Relative dates. Google shows 3 days ago, not the exact date, so postedAt is accurate to the hour for recent jobs and to the day for older ones.
• Currency. When Google omits the currency symbol, the selected country's currency is used and salaryCurrencyInferred is true.
• Location matters. Without a location, Google picks an area inside the country, which can vary between runs. Set a location for reproducible results and stable monitoring.
• Filter checks. Jobs that show no date or type cannot be verified and are kept, because Google included them in the filtered results.
• Staffing agency list is a documented heuristic: well-known agencies (Robert Half, Randstad, Adecco, Manpower, Kelly Services, Kforce, Insight Global, TEKsystems, Aerotek, Actalent, Aston Carter, Apex Systems, Hays, Michael Page, Robert Walters, CyberCoders, Jobot, Experis, AMN Healthcare, Aya Healthcare, Cross Country, Medical Solutions, Vivian Health and others) plus names containing Staffing, Recruiting, Recruitment, Workforce Solutions or Servicios Temporales. Use Exclude companies for anything else.
• Monitoring memory keeps up to 20,000 jobs per query; the oldest are forgotten first.

FAQ

Is it legal? The Actor collects publicly available job postings published by companies. It does not log in and does not collect personal data about candidates. Make sure your use complies with the laws that apply to you.

How do I get only new jobs every day? Enable Only new jobs since last run, set Date posted to "Last 3 days" and schedule the task daily. Each run saves and charges only jobs not delivered before.

Why did I get fewer jobs than "Max jobs per query"? Check stopReason in RUN_SUMMARY. Most often Google has no more unique listings for that search; try a broader query or nearby locations.

Do I need my own proxy? No. The Google SERP proxy is included in the price.

Can I scrape the company's own career site instead? Yes, with ATS Jobs Scraper by the same publisher, which reads Greenhouse, Lever, Ashby and Workday boards directly and also reports closed jobs.

Related Actors from ocius-data

Built to the same standard: a status for every input in RUN_SUMMARY, actionable error messages and list-based inputs, so they chain well in one workflow:

• ATS Jobs API: Greenhouse, Lever, Ashby, Workday: jobs straight from company career sites on Greenhouse, Lever, Ashby and Workday, including closed jobs.

Changelog

• November 2026 New keep-only filters: Only these companies, Only jobs published on (for example LinkedIn or Indeed) and Only jobs with salary. Filtered jobs are not charged.
• 0.2 (November 2026) Monitoring mode (only new jobs since the last run, per query and settings, status NO_NEW_JOBS); Exclude companies and Exclude staffing agencies filters; README with integrations, cost calculator and field dictionary.
• 0.1 (November 2026) First version: Google Jobs search with location, country and language; date and employment type filters with double verification; refinement fan-out beyond 10 results; deduplication; salary, date and type normalization; all apply links; per-query statuses and run summary.

Support

Open an issue on the Actor page. Issues are reviewed every day and answered within one business day.