Lever Jobs Scraper - export jobs.lever.co boards to CSV, JSON & Excel

Scrape live job postings from any Lever job board (jobs.lever.co) and get clean, flat rows you can open in Excel, load into a sheet, or pipe into your own tooling. Give the Actor a list of Lever company slugs - or just paste the board URLs - and it returns every public posting on those boards with Lever's full category block normalized: team, department, commitment, workplace type (remote / hybrid / onsite), location, all locations and country, plus the full job description as readable plain text.

It talks to Lever's own public postings API (https://api.lever.co/v0/postings/{company}?mode=json), so there is no API key, no login and no browser involved. A whole board comes back in one request with no pagination, so keeping a job feed, an ATS dataset or a recruiting CRM up to date costs one request per company. Turn on Only postings new since the last run and a scheduled run gives you just the new roles.

This Actor covers Lever only, on purpose: one platform, done properly. If you need several ATS platforms at once (Greenhouse, Ashby, SmartRecruiters, Recruitee, Workable), use our ATS Job Postings Scraper.

What you get

One row per job posting.

Field	Type	Description
company	string	The Lever company slug the posting came from.
jobId	string	Lever's posting ID (a UUID). Stable - good as a primary key.
title	string	Job title, e.g. Senior Product Manager - User Platform.
team	string	Lever category team, e.g. Platform. Null when the employer leaves it blank.
department	string	Lever category department, e.g. Engineering.
commitment	string	Lever category commitment, e.g. Full-time, Permanent, Intern.
workplaceType	string	remote, hybrid or onsite, as published by the employer. Null if not set.
location	string	Primary location, e.g. London.
allLocations	string	Every location on the posting, ; separated, e.g. London; Stockholm.
country	string	Two-letter ISO country code, e.g. GB, US.
descriptionText	string	The job description as plain text (up to 20,000 characters).
listsText	string	Lever's structured sections ("What You'll Do", "Responsibilities", "Who You Are"...) flattened into headed bullet lists.
hostedUrl	string	Public posting page on jobs.lever.co.
applyUrl	string	Direct link to the application form.
createdAt	ISO 8601	When the posting was created (Lever returns epoch milliseconds; converted here).
scrapedAt	ISO 8601	When this row was collected.

A SUMMARY record is written to the run's key-value store with the companies requested, which ones succeeded, which returned nothing, the department / workplace-type / country breakdowns, and whether the run stopped on its time limit.

Input

Field	Type	Default	Description
companies	array of strings	["spotify", "wealthfront", "palantir"]	Lever company slugs or full board URLs. spotify, https://jobs.lever.co/spotify and https://jobs.lever.co/spotify/<job-id> all resolve to the same board.
maxResultsPerCompany	integer	200	Cap on rows saved per company. Postings are sorted newest first, so a low cap keeps the freshest roles.
onlyNewSinceLastRun	boolean	false	Save only postings not seen in an earlier run. IDs are kept in a named key-value store (lever-jobs-baseline). The first run per company records the baseline and returns everything.
maxRunSeconds	integer	240	Wall-clock budget. On reaching it the Actor stops cleanly, keeps what it saved and logs what it skipped, instead of dying on a timeout.

The slug is the path segment in https://jobs.lever.co/<slug>. A slug with no Lever board returns HTTP 404; that company is reported in the log and in SUMMARY and the run continues with the rest. If every company fails, the run fails loudly with the reason rather than handing you an empty dataset.

Example output

A real row from a default run (description and section text truncated here for readability):

{
  "company": "wealthfront",
  "jobId": "4d469c84-f50e-479f-bc29-80b573cd6336",
  "title": "Market Commentary Writer + Social Editor",
  "team": "Communications",
  "department": "Communications",
  "commitment": "Full-time",
  "workplaceType": "remote",
  "location": "San Francisco, CA",
  "allLocations": "San Francisco, CA; New York City, NY",
  "country": "US",
  "descriptionText": "Opportunity:\nWealthfront is looking for an experienced financial writer and/or editor (or audio reporter/producer with writing experience) to support our growing market commentary vertical and newsletter... [truncated here]",
  "listsText": "Responsibilities:\n- Monitor market and macroeconomic news daily and generate high-quality, timely content ideas that help young professionals navigate market trends and build their financial confidence... [truncated here]",
  "hostedUrl": "https://jobs.lever.co/wealthfront/4d469c84-f50e-479f-bc29-80b573cd6336",
  "applyUrl": "https://jobs.lever.co/wealthfront/4d469c84-f50e-479f-bc29-80b573cd6336/apply",
  "createdAt": "2026-09-16T17:42:52.563Z",
  "scrapedAt": "2026-09-24T00:50:57.581Z"
}

FAQ

How do I find a company's Lever slug? Open the company's careers page. If the postings live on jobs.lever.co/<something>, that <something> is the slug. You can paste the whole URL into companies - the Actor extracts the slug itself.

The company I want returns "no Lever board". What now? Not every employer uses Lever, and a 404 from api.lever.co means exactly that: there is no Lever board under that slug. Check the careers page to see which ATS the employer actually uses. For example, Thumbtack's public board is on Ashby, not Lever (checked 2026-09-23) - our ATS Job Postings Scraper covers Ashby, Greenhouse, SmartRecruiters, Recruitee and Workable.

A company responded but returned zero postings. Is that a bug? No. Lever answers 200 with an empty list when the board exists but has nothing public right now. That case is logged and listed in SUMMARY under companiesWithNoPostings, so it is easy to tell apart from a wrong slug.

How do I get only new jobs on a schedule? Set onlyNewSinceLastRun to true and schedule the Actor. The first run records every posting ID it sees and returns them all; later runs return only postings that were not in the baseline. A run with nothing new finishes successfully with an empty dataset.

How fast is it, and how many companies can I pass? One HTTP request per company, no pagination, plus the one-second crawl delay that api.lever.co/robots.txt publishes. On 2026-09-23 the default three-company run saved 298 postings end to end in 4 to 7 seconds across repeated runs (Lever's own response time is the variable part). Board sizes change day to day, so treat that as the shape of it rather than a fixed number. Budget roughly a second per extra company, and raise maxRunSeconds (and the run timeout in Run options) for long lists so the run is not cut short.

Does it return salary data? Only if the employer publishes it in the description text - Lever's public postings API does not expose a structured salary field on these boards, so this Actor does not invent one.

Data source / access / limits

• Source: Lever's public postings API, https://api.lever.co/v0/postings/{company}?mode=json. This is the same feed that renders a company's public jobs.lever.co board.
• Access: no API key, no account and no cookies. Only postings the employer has made public are returned; nothing behind a login is touched.
• Coverage: Lever boards only. Companies on other ATS platforms are out of scope for this Actor by design.
• Freshness: whatever the board shows at the moment of the run. createdAt is the posting's creation time from Lever; scrapedAt is when the row was collected.
• Rate limits and robots: api.lever.co/robots.txt allows this path and publishes Crawl-delay: 1, which the Actor honours by waiting a second between companies. Lever documents no request quota for this endpoint. The Actor sends one request per company and retries 429 and 5xx responses with exponential backoff.
• Billing: this Actor is paid per result on the Apify platform - you are charged for the rows it saves. The maxResultsPerCompany cap and onlyNewSinceLastRun both exist so you can keep that number where you want it.
• Your responsibility: job postings are published by employers for public consumption, but how you store and use the data (including any personal data in a description) is up to you and your jurisdiction.