Lever Jobs Scraper: One Board You Name, No API Key
Pricing
from $1.50 / 1,000 job postings
Lever Jobs Scraper: One Board You Name, No API Key
Give it a Lever board token or careers URL and get that company's open roles, live from the source. No key, no login, no cookies.
Pricing
from $1.50 / 1,000 job postings
Rating
0.0
(0)
Developer
glitchbound
Maintained by CommunityActor stats
0
Bookmarked
1
Total users
0
Monthly active users
5 hours ago
Last modified
Categories
Features
Share
Lever Jobs Scraper: Any Board, No API Key
Every open role from any Lever job board, by company domain or board token. Titles, locations, teams, commitment type and full descriptions as clean JSON. No API key and no proxy: give it a domain and it finds the board itself.
Features
- Any lever board, from the URL or token you already have. No API key, no login, no cookies, no proxy: this reads the public feed lever publishes so that job boards can index it.
- Every posting, not the first page. Paging is handled, including the places where lever reports a total it does not honour.
- One row per job, as clean JSON, with the same field names on every run.
- Filters that cost you nothing. Rows dropped by
titleKeywords,titleExclude,locationKeywordsorremoteOnlyare never charged for. - Errors are per company. One bad id does not end the run, and error rows are not charged.
- Run it on a schedule and the rows become a record of who started hiring and when.
How to use it
- Click Try for free, or add this Actor to a task.
- Put one or more companies or board ids in the input. The example below is a real one.
- Optionally narrow it with
titleKeywords,locationKeywordsorremoteOnly. Filtered rows are not billed. - Run it. Results appear in the dataset and can be exported as JSON, CSV, Excel or fetched from the API.
Input
| Input | Type | Default | What it does |
|---|---|---|---|
companies | array of strings | none | Company domains or names, one per line, e.g. palantir.com |
boards | array of strings | none | Board tokens to read, one per line, e.g. palantir |
titleKeywords | array of strings | none | Keep only roles whose title contains any of these, one per line, e.g. security or staff engineer |
titleExclude | array of strings | none | Drop roles whose title contains any of these words. Whole words (intern keeps International); end a word with * for a prefix; senior also drops Sr. Never charged. |
locationKeywords | array of strings | none | Keep only roles whose location contains any of these, one per line, e.g. London or Israel |
remoteOnly | true or false | false | Keep only roles flagged remote by the ATS, or whose title or location says remote or anywhere |
includeDescription | true or false | false | Fetch the description text as well |
maxResultsPerCompany | number | 1000 | Ceiling on roles taken from any single board |
maxItems | number | none | A hard ceiling on rows for the entire run, across every company |
{"boards": ["palantir"],"titleKeywords": ["engineer"],"remoteOnly": true}
Output
One row per job. This is the shape, with the fields lever actually publishes:
{"title": "Senior Software Engineer","location": "Berlin, Germany","url": "https://boards.example.com/jobs/8130725","jobId": "8130725","postedAt": "2026-08-20T09:14:02Z","departments": ["Engineering"],"offices": ["Berlin"],"description": "We are looking for...","employmentType": "full-time","isRemote": true,"company": "palantir","atsPlatform": "lever","boardToken": "palantir","boardUrl": "https://boards.example.com/palantir","seniority": "senior"}
| Field | |
|---|---|
title | the role as the company wrote it |
location | as the board published it, EXCEPT where the board published a count instead of a place. Workday writes "2 Locations" into this field on multi-office postings, and 27% of some boards' rows carried a tally where a place belonged; those now carry the primary office read from the job's own URL, and locationCount keeps the number |
url | the public posting, ready to open |
jobId | the ATS's own id, stable across runs |
postedAt | ISO 8601 UTC |
departments | list |
offices | list |
description | opt in, off by default because it is expensive at the source |
salaryMin, salaryMax, salaryCurrency, salaryPeriod, compensation | the pay band, from Lever's own pay field or the posting's text, where the employer states one |
employmentType | full time, contract, intern, as the source says |
isRemote | true only for genuinely remote roles |
company | what you asked for, echoed back |
atsPlatform | which system it came from |
boardToken | the board id used |
boardUrl | the public board this came from |
city | where the location names one |
country | derived from the location, and from the job URL where the location is a phrase like "2 Locations" |
region | state or province, where the location names one |
seniority | derived from the title: intern, junior, senior, staff, lead, principal, director or executive. Empty where the title does not say, which is most of them |
workArrangement | remote, hybrid or onsite |
locationCount | how many offices one posting is open in, present only where the board published a count instead of a place |
applicationDeadlineText | "3 days left to apply", on the few postings where the board publishes a closing date INSTEAD of a posting date. On those rows postedAtRelative is absent, because a deadline is not a posting time |
What people use this for
Hiring data is not really about jobs. It is the earliest public signal a company gives that something changed, and it is why three different kinds of buyer end up on the same dataset:
- Sales and go-to-market. A company that opens six engineering roles this month is a company with new budget. Job postings say which team is growing and in which city, weeks before anything shows up in a funding announcement.
- Investors and market research. Headcount by function, tracked over time, across a whole portfolio or a whole sector. Every row carries the company, the team and the date, so a weekly run is a time series.
- Recruiting and talent. Where a competitor is hiring, which roles they have been trying to fill for months, and how fast a team is growing. Where the employer fills in Lever's own pay field (
salaryRange, on 54 of the 60 largest US Lever boards), it comes back assalaryMin,salaryMax,salaryCurrencyandsalaryPeriod; where it writes the band into the posting instead, as Palantir does ("$135,000 - $200,000/year"), that is read the same way. Lever sends the whole posting with every job, so neither needsincludeDescription. A posting that states none gets none.
Run it once for a snapshot. Run it on a schedule and the same rows become a record of who started growing and when.
You do not need the board token
{ "companies": ["palantir.com"] }
Every other lever Actor asks you for a board token, and there is no directory
mapping a company to its token. This one derives candidates from the domain and
verifies each against the live API, which is sound because lever answers
404 for a board that does not exist and 200 for one that exists with nothing
open. A 200 is proof, not a maybe.
Measured across 30 companies, deriving and verifying resolved 90% against 50% for reading the careers page for a link. Stripe's careers page is a JavaScript app that names no board at all, and it still resolves.
What makes lever different
-
Measured:
palantir302 open roles,spotify105,walkme53. -
Lever answers with a bare JSON array, not an object with a
jobskey like every other system here. A client written for Greenhouse reads that as zero results. -
Lever carries
commitment(full time, intern, contract) and the posting in plain text.descriptionis the whole of it as the job page shows it: the description, the requirement lists and the closing section, which is where US employers state the pay.
Filters that cost you nothing
Filtered rows are not charged. Keywords match as plain text, so c++ and
node.js mean exactly that rather than being read as regular expressions.
Integrations and API
Every run writes to a dataset you can export as JSON, CSV or Excel, or read from the Apify API. The Actor can be scheduled, called from another Actor, or wired into Make, Zapier, Slack, Google Sheets and the rest of Apify's integrations. It is also callable by an AI agent through the Apify MCP server, and the output schema means the agent gets field descriptions rather than raw JSON.
Looking for an Indeed or Glassdoor API? There isn't one, and this is why you don't need it
Indeed has no public jobs API. Neither does Glassdoor, ZipRecruiter or LinkedIn Jobs, and all four block you at the edge. Measured from an ordinary residential address on 2026-08-01, with normal browser headers:
indeed.com/jobs 403 0 bytesglassdoor.com/Job/... 403 0 bytesziprecruiter.com 403 0 bytesupwork.com/nx/search 403 0 bytes
Zero bytes. Cloudflare rejects the request before it reaches an application, so there is nothing to parse and no proxy budget that fixes it.
But none of those four originate job data. They aggregate it from company career pages, and those pages run on systems like lever that publish a free, keyless, public API, because companies want their openings indexed. That API answered with real jobs from the same connection, in the same minute.
Going to the source is also fresher. An aggregator shows you its last crawl. This shows you the board.
Leave titles out. titleExclude drops roles whose title holds any of its words, such as senior or intern. Whole words, so intern keeps International; end a word with * for every word it begins (intern* also drops Internship), and senior also drops Sr. Dropped roles are never returned and never charged.
New since your last run. Every job says whether an earlier run of the same careers site already sent it to you: newSinceLastRun (true or false), firstSeenAt (the day you first got it) and previousRunAt. Schedule the run daily and the new jobs are the rows marked true, with nothing to diff yourself. The history stays in a key-value store in your own account (lever-jobs-seen); a site's first run carries firstSeenAt only. It changes no row count and no price. Set onlyNewJobs to get only those jobs and pay only for them: the first run records the starting point and returns nothing, with a note saying so. New means new to you: a job past an earlier run's limits or filters counts as new the first time it is sent.
FAQ
Do I need an API key or an account with lever? No. This reads the public feed lever publishes for indexing. Nothing here is behind a login, a paywall or a bot wall.
What am I charged for? Rows returned. Rows removed by your filters are not charged, and neither are error rows.
Can I get only what changed since my last run?
Yes. With trackNewJobs on (the default), each job says whether an earlier run
of that careers site already sent it to you: newSinceLastRun, firstSeenAt
and previousRunAt, from a history kept in your own Apify account. Set
onlyNewJobs to get, and pay for, only the new ones; a site's first run records
the starting point. Schedule it daily and each run is that day's new jobs.
A board came back empty. Is the company not hiring? An empty board is a real answer here and is reported as one.
Notes
- The source is a public API that companies publish deliberately. Nothing here is behind a login, a paywall or a bot wall.
- An empty board is a real answer and is reported as one.
- Errors are per company. One bad token does not end the run, and error rows are not charged.
Related
- Company Career Page Jobs Scraper: any company's jobs by name or domain, on 24 career systems
- Workday Jobs Scraper: Workday career sites by company name
- Job Listings API: search every company at once, with a daily feed of new jobs
Use it from an AI agent
This Actor is a ready-made tool for Claude, ChatGPT, Cursor or any other MCP client, through Apify's MCP server. Add it as a connector at:
https://mcp.apify.com?tools=glitchbound/lever-jobs-scraper
In Cursor or VS Code the same address goes in mcp.json:
{ "mcpServers": { "apify": { "url": "https://mcp.apify.com?tools=glitchbound/lever-jobs-scraper" } } }
The first call asks you to sign in to Apify, or takes an API token. From then on the agent sees this Actor's input fields and calls it directly.
Feedback
If this saved you time, a review on the Actor's Store page helps other people find it. If something is missing or wrong, the Issues tab on the same page is the fastest way to reach me, and it gets read.