Glassdoor Jobs Scraper

Scrape Glassdoor job listings with full job descriptions, salary ranges that tell you where each figure came from, and no charge for duplicate results.

$1 per 1,000 jobs. Full descriptions are an optional add-on at $5 per 1,000.

What you get

One row per job. Every field is present on every row, with null where Glassdoor has no value — so your downstream code can rely on the shape.

jobId · title · companyName · companyRating · location · isRemote · salaryMin · salaryMax · salaryCurrency · salaryPeriod · salarySource · isEasyApply · postedAt · jobUrl · companyUrl · descriptionSnippet · descriptionText · descriptionHtml · scrapedAt

Measured across a 1,000-job run: 100% coverage on titles, companies, locations, URLs and posting dates. 94.7% on company rating. 91.7% on salary.

Full job descriptions

Switch on fetchFullDescriptions and every row carries the complete posting text — about 4,900 characters on average, up to 18,000 — as both plain text and HTML.

Glassdoor's search results carry only a ~120-character preview. The full text lives on each job's own page, which means a separate fetch per job. So you get the preview in descriptionSnippet and the full text in descriptionText, and descriptionText stays null when you haven't asked for it. A truncated description never masquerades as a complete one.

Because each description is its own page fetch, it is priced separately and off by default. A 1,000-job run with descriptions takes about an hour.

Salary you can check

salarySource tells you whether each figure came from the employer or from Glassdoor's own estimate. In a 1,000-job sample: 548 rows employer-provided, 369 estimated.

Salary analysis built on unlabelled data is analysis built on a mix of stated facts and models, with no way to tell them apart. This field is how you tell them apart.

You are not charged for duplicates

Glassdoor recycles listings across pages and across searches. A typical run collects 200-250 repeats. This scraper drops them and does not bill you for them.

Charges fire per unique job written to your dataset — never per request, never for a page that failed, never for a row that came back empty. maxItems is a hard ceiling: the run stops there and stops charging there.

How many jobs one search returns

Glassdoor stops serving new results for a single search after roughly 30-90 jobs, then repeats listings it has already shown. That is Glassdoor's limit, not this scraper's, and every Glassdoor scraper meets it.

To collect more, pass several entries in searchQueries. Narrower titles and separate locations each get their own window of results, and duplicates across them are removed automatically. A 1,000-job run typically needs 30-40 searches.

If a run falls short, it tells you how many pages it read, how many unique jobs it found, and why it stopped. It does not truncate silently.

Locations

Cities, states and whole countries. Pass a place name in location and it is resolved against Glassdoor's own location index rather than a built-in list, so there is no fixed set of supported countries.

Verified in the United States, United Kingdom, Germany, Canada, India, the Netherlands and Australia.

Pricing

Event	Price
Job scraped	$0.001 — $1 per 1,000
Full description fetched	$0.005 — $5 per 1,000

Each run costs $0.00005 to start. Beyond that you pay only for unique jobs saved to your dataset — duplicates, failed pages and blocked requests are free.

Maintenance

Glassdoor changes its site, and when it does, scrapers break. This one is maintained: report a problem through the Issues tab and it gets fixed.