indeed-scraper

Scrape Indeed job listings into clean, typed JSON or CSV: structured salary, company rating, location, job type and posting date, with no CSS selectors to maintain.

Search Indeed for any job title, skill or company in any location and get back one clean record per job: job key, title, company and its Indeed rating, city / state / postal code, remote or hybrid setting, salary as text and as min / max / pay period / currency, job types, the date Indeed published it, sponsored / urgently-hiring / "easily apply" flags, a snippet and the job URL. Add --details to also get the full description, benefits and company logo. Tested on Indeed's US, UK and Canada sites; other country sites are addressed the same way.

Powered by ScrapeUnblocker. Every Indeed page is loaded in a real browser through the getPageSource API, and jobs are read from the structured data Indeed embeds in each results page, not from the rendered cards.

Features

• 💰 Salary you can sort by: "$152,513 - $190,642 a year" becomes salary_min: 152513, salary_max: 190642, salary_period: "year", salary_currency: "USD" (hourly, daily, weekly and monthly pay too; GBP on the UK site, CAD on the Canadian one).
• 🎛️ Indeed's own filters: date posted (last 1 / 3 / 7 / 14 days), job type (full-time, part-time, contract, temporary, internship), remote only, search radius, and sort by relevance or date.
• 🎯 Direct matches only: Indeed mixes sponsored "related jobs" groups, nearby-location jobs and recommendations into its results. Every job carries a match_type, and by default you only get the jobs that actually match your search (--include-related keeps the rest).
• 🔎 Many searches at once: pass several queries and locations; every query runs in every location and the results are merged and de-duplicated by job key.
• 📄 Full postings on demand: --details adds the complete description (as plain text), benefits and company logo for every job; scraper.job_details() also returns the description HTML.
• 📅 Real dates: date_posted is Indeed's publication date (2026-09-17), not just "10 days ago" (kept too, as posted_ago).
• 🌍 Country sites: --country GB searches uk.indeed.com, CA ca.indeed.com, and so on; URLs in the output point at the right site.
• 🔁 Resilient: bot-check pages and timeouts are retried with doubling backoff, and a search with no matches comes back as an empty list.
• 🖥️ CLI + library: use indeed-scraper from the shell or import it in Python.

Install

pip install .
# or, for development:
pip install -e ".[dev]"

Set your ScrapeUnblocker API key (grab one from your dashboard):

cp .env.example .env          # then edit it, or just export the variable:
export SCRAPEUNBLOCKER_KEY=your_key_here

CLI usage

# One search, pretty-printed JSON to stdout
indeed-scraper "python developer" -l Remote --pretty

# Two roles in two cities, posted in the last 3 days, newest first, as CSV
indeed-scraper "data analyst" "data engineer" -l "New York, NY" -l "Austin, TX" \
  --days 3 --sort date --format csv --output data_jobs.csv

# Full-time nursing jobs on Indeed Canada
indeed-scraper "registered nurse" -l "Toronto, ON" --country CA --job-type fulltime

# Remote DevOps jobs with full descriptions and benefits (1 extra request per job)
indeed-scraper "devops engineer" --remote --details --limit 10 --output devops.json

# UK jobs within 10 miles of Manchester, without sponsored listings
indeed-scraper "accountant" -l Manchester --country GB --radius 10 --no-sponsored

# Any search you set up on indeed.com: just paste the URL
indeed-scraper --url "https://www.indeed.com/jobs?q=product+manager&l=Seattle%2C+WA&fromage=7"

Progress goes to stderr (here for indeed-scraper "python developer" -l Remote --limit 3 --details -o jobs.json):

[indeed] "python developer" in Remote (US): 15 direct matches, 26 related/sponsored cards (2,430 matching)
[indeed] details a80654465f93f95d: ok
[indeed] details 9e465570cba1c5ce: ok
[indeed] details 86aa76a7dde09ebc: ok
Wrote 3 jobs to jobs.json

usage: indeed-scraper [-h] [-l PLACE] [--url URL] [-c ISO]
                      [-s {relevance,date}] [-d {1,3,7,14}]
                      [-t {fulltime,parttime,contract,temporary,internship}]
                      [--remote] [-r DIST] [--no-sponsored]
                      [--include-related] [-n LIMIT] [--details]
                      [--proxy-country ISO] [--retries N] [-f {json,csv}]
                      [-o PATH] [--pretty] [-q] [--version]
                      [QUERY ...]

Option	Meaning
QUERY ...	Job title, skill or company. Several queries are merged and de-duplicated.
-l, --location	City, state, postal code or Remote (repeatable; every query runs in every location).
--url	A full Indeed search URL instead of QUERY / --location (repeatable).
-c, --country	Which Indeed site: US (default), GB, CA, AU, IN, DE, …
-s, --sort	relevance (Indeed's default) or date (newest first).
-d, --days	Only jobs posted in the last 1, 3, 7 or 14 days.
-t, --job-type	fulltime, parttime, contract, temporary or internship.
--remote	Only remote jobs (Indeed's "Remote" filter).
-r, --radius	Search radius around the location (miles on indeed.com).
--no-sponsored	Drop sponsored jobs.
--include-related	Also keep Indeed's related-job groups, nearby jobs and recommendations.
-n, --limit	Stop after this many unique jobs.
--details	Fetch each job's full posting (description, benefits, logo). One request per job.
--proxy-country	Country to route requests through (default US, which works for every Indeed site).
--retries	Retries per request on a transient error or bot-check page (default 3, doubling backoff).
-f, --format / -o, --output	json (default) or csv (needs --output).

Library usage

from indeed_scraper import IndeedScraper, to_csv

scraper = IndeedScraper()  # reads SCRAPEUNBLOCKER_KEY

jobs = scraper.search("data analyst", "New York, NY", days=7, sort="date")
for job in jobs:
    if job.salary_min is not None:
        print(f"{job.title} @ {job.company}: {job.salary_min}-{job.salary_max} "
              f"{job.salary_currency} per {job.salary_period}")

# Every query in every location, merged and de-duplicated by job key
more = scraper.search_many(
    ["data analyst", "data engineer"], ["Chicago, IL", "Remote"], job_type="fulltime"
)

# Full postings (description, benefits, logo), one request per job
scraper.enrich(more[:5])

to_csv(jobs + more, "indeed_jobs.csv")

Other entry points:

• scraper.search("nurse", "Toronto, ON", country="CA") searches another Indeed site.
• scraper.scrape_url(url) scrapes any Indeed search URL, and scraper.scrape_many([url1, url2]) merges several.
• scraper.job_details("9e465570cba1c5ce") returns one job's full posting as a JobDetails (or None if Indeed no longer opens it).
• build_search_url("nurse", "Chicago, IL", days=3, job_type="parttime") just builds the URL.
• on_page=callback receives each SearchPage, with Indeed's total match count, the echoed query / location, and every card on the page whatever its match_type.
• IndeedScraper(retries=..., backoff=..., on_retry=callback) tunes the retry policy.

Example output

A real job from indeed-scraper "python developer" -l Remote --details (description and benefits trimmed):

{
  "job_key": "9e465570cba1c5ce",
  "title": "Python Developer",
  "company": "DMI",
  "company_url": "https://www.indeed.com/cmp/Dmi-a4790562",
  "company_rating": 3.6,
  "company_review_count": 171,
  "location": "Remote",
  "city": null,
  "state": null,
  "postal_code": null,
  "country": "US",
  "work_setting": "Hybrid work",
  "remote": true,
  "salary_text": "$152,513 - $190,642 a year",
  "salary_min": 152513,
  "salary_max": 190642,
  "salary_period": "year",
  "salary_currency": "USD",
  "job_types": ["Full-time"],
  "date_posted": "2026-09-17",
  "posted_ago": "10 days ago",
  "sponsored": false,
  "urgently_hiring": false,
  "easy_apply": false,
  "match_type": "search",
  "related_group": null,
  "snippet": "The program provides IT support to thousands of users across Washington, DC and nationwide field locations. This role works directly with program leadership and…",
  "url": "https://www.indeed.com/viewjob?jk=9e465570cba1c5ce",
  "search_query": "python developer",
  "search_location": "Remote",
  "description": "About DMI:\n\nDMI is a leading provider of digital services and technology solutions, headquartered in Tysons Corner, VA. …",
  "benefits": ["Referral program", "Tuition reimbursement", "Pet insurance", "Health insurance", "401(k) matching"],
  "company_logo": "https://d2q79iu7y748jz.cloudfront.net/s/_squarelogo/256x256/7bebb0154343687c6caf0c3ad192e392"
}

Field notes:

• salary_min / salary_max are per salary_period (hour, day, week, month or year). "From $80,000 a year" gives salary_max: null. The figure is the one Indeed shows: posted by the employer or extracted by Indeed from the description. Jobs without a salary have all salary fields null.
• remote is true when Indeed lists the job as remote; work_setting is Indeed's work-setting label (Remote, Hybrid work, …) and can be more specific, as above.
• date_posted is Indeed's original publication date. Postings that stay open keep it, so a job shown as "30+ days ago" can date back months.
• match_type is search for a direct match. With --include-related you also get related (Indeed's sponsored "Jobs in …" groups, named in related_group), nearby (sponsored jobs from nearby locations) and recommended (Indeed's suggestions when a search matches little or nothing).
• company_rating / company_review_count are the company's Indeed reviews (null when it has none). easy_apply means the job takes Indeed Apply.
• description, benefits and company_logo are only filled with --details (JobDetails also carries description_html).
• Missing values are null, never empty strings. In CSV output, job_types and benefits are joined with " | ".

Good to know

• One results page per search. Indeed shows signed-out visitors the first page of results only and asks them to sign in for page 2, so a search returns up to about 15 direct matches (Indeed's total count is still reported). To collect more, split the search: several queries, several locations, a date-posted window, --sort date next to relevance. search_many (or several QUERY / -l values on the CLI) merges and de-duplicates for you.
• Details cost a request each. --details opens every job's full posting, so 15 jobs with details is 16 requests. Use --limit to cap it.
• Country sites. Tested on the US, UK and Canada sites; other Indeed country sites (au.indeed.com, in.indeed.com, de.indeed.com, …) are addressed the same way. Requests go through a US route by default, which returns the same local listings as a local route on those sites; --proxy-country changes it.
• Retries. Timeouts and bot-check pages are retried with doubling backoff (3, 6, 12 seconds by default). Each attempt is a normal API request.
• No results. A search with no matches returns an empty list, not Indeed's "recommended jobs" filler.

Project layout

src/indeed_scraper/
  __init__.py      package exports + version
  scraper.py       Job / JobDetails models, IndeedScraper, URL builder, page parsing, JSON/CSV export
  cli.py           argparse command-line interface
examples/          runnable scripts (search → JSON, multi-search → CSV, daily new-jobs alert)
tests/             offline unit tests (SDK mocked, no API credit spent)

Development

make install       # editable install with dev deps
make lint          # ruff check
make format        # ruff format
make test          # pytest (offline; the SDK is mocked)
make run           # run the search → JSON example (needs SCRAPEUNBLOCKER_KEY)

The test suite mocks the ScrapeUnblocker client and uses synthetic Indeed pages, so make test runs fully offline and spends no API credits.

Links

• 🌐 ScrapeUnblocker
• 📚 Documentation
• 🐍 Python SDK

Legal

For research and educational use. Respect Indeed's Terms of Service and robots.txt, data-protection law (job postings can name recruiters and contacts) and other applicable laws when scraping and storing job data. This project is not affiliated with or endorsed by Indeed.

License

MIT, see LICENSE. Copyright (c) 2026 ScrapeUnblocker.