Remote Jobs Aggregator — 8 job boards, one clean feed

Pull remote job listings from 8 public job boards in a single run, normalized to one schema, deduplicated across sources, and filtered by keyword, location, recency and salary — as clean JSON, CSV or Excel.

• One schema, 8 sources. Every board's quirks (integer vs. free-text salary, RSS vs. JSON, remote flags, restricted geos) are normalized into the same flat record, so you never write per-source parsing code.
• Deduplicated. The same opening posted on more than one board is merged into a single record; the richest copy is kept (salary data > longer description > newest) and alsoSeenOn lists every other board it appeared on.
• No login, no API key, no browser. Plain HTTP against each board's public API/feed. Fast and cheap — a 300-job run across all 8 sources typically finishes in under a minute.
• Pay only for results. Filtered-out jobs and merged duplicates are free; you are charged only for the records actually saved.

Sources

Source	Type	Notes
RemoteOK	JSON API	Fully remote-only board; salary as numbers when published
Remotive	JSON API	Opt-in (off by default) because of its redistribution terms. Free tier caps at ~15 jobs per call (verified live)
Himalayas	JSON API, paginated	Numeric salary, seniority and category fields; largest volume of the 8
Jobicy	JSON API	Up to 200 jobs per call (server-side cap)
Arbeitnow	JSON API, paginated	Mixed on-site/remote board; only remote: true postings are kept
We Work Remotely	RSS feed	Title format Company: Role, split automatically
Working Nomads	JSON API
The Muse	JSON API, paginated	Filtered to location = Flexible / Remote

What can you do with it?

Use case	How
Build or seed a niche job board	keywords: ["python", "data engineer"], dedupe: true for one clean, deduplicated feed across every source
Recruiter / sourcing lead lists	keywords, locations, postedWithinDays: 7 for a fresh weekly digest of who's hiring for a role
Hiring-signal / market research	Run on a schedule, track which companies post remote roles and how salary bands move over time
Job-alert automations	Schedule the run, pipe new records into Slack/Telegram/email via Make, n8n or Zapier
AI agents	Simple input, flat JSON output, works through the Apify MCP server
Daily feed of only new remote jobs	onlyNew: true on a daily schedule: every run delivers just the jobs you have not received before, and you pay only for those
Watch specific employers	companyBoards: ["greenhouse:stripe", "ashby:linear", "lever:spotify"] to follow the remote openings of companies you care about

Input

Field	Example	Notes
sources	["remoteok", "remotive"]	Which boards to query. Leave empty for the 7 default boards (Remotive is opt-in, see Sources).
keywords	["python", "data engineer"]	Keep a job if ANY keyword matches title/company/tags (case-insensitive)
excludeKeywords	["senior"]	Drop a job if ANY of these matches
titleOnly	true	Match keywords against the title only
locations	["Europe", "Worldwide"]	Substring match against the job's location text; "Worldwide" also matches jobs with no region restriction
postedWithinDays	14	Only jobs posted in the last N days. Jobs with an unknown post date are always kept
minSalaryUsd	100000	Only applied to jobs with USD salary data; non-USD or missing salary never gets filtered out. Hourly/monthly figures are annualized
includeDescription, descriptionFormat	true, "text"	text strips HTML to clean readable text; html keeps the source's raw HTML
dedupe	true	Merge the same job across sources
maxItemsPerSource	500	Cap on raw jobs scanned per source before filtering (for company boards: cap on the roles kept after remoteOnly)
maxItems	200	Hard cap on saved (and charged) jobs
onlyNew	true	Deliver only jobs not returned by an earlier run with the same stateKey (default false)
stateKey	"python-jobs"	Name of the "already seen" list (default "default")
companyBoards	["greenhouse:stripe", "https://jobs.lever.co/spotify"]	Extra company career boards (Greenhouse, Ashby, Lever)
remoteOnly	true	For company boards only: keep just roles advertised as remote (default true)

Only new jobs (scheduled runs)

Turn on onlyNew and schedule the Actor (for example daily). The first run returns everything that matches and remembers it; every later run returns only jobs that were not delivered before, so your downstream Slack/Sheets/CRM flow never sees the same job twice. Jobs that are skipped as already seen are free, you are charged only for the new records saved.

How it works:

• The list of seen jobs lives in a named key-value store remote-jobs-state in your Apify account, under the record seen-<stateKey>. Each entry is one job posting (the board's own job id, or the posting URL) with its first-seen and last-seen dates. It is only updated after the jobs were successfully saved, and jobs cut off by maxItems or your max cost are not marked as seen, so they arrive on the next run.
• "New" means a posting you have not been given. A job you already received that later shows up on one more board is not new (with dedupe on it is merged into the record you already have). A genuinely new posting of a role you already have — say the same title opened for a second location on the company's Greenhouse board — is new and is delivered.
• A posting is refreshed every time it is still listed, so a job that stays open for months is never delivered twice. Postings not seen for 90 days are forgotten (and would count as new if they came back), and the list is capped at 200,000 postings (least recently seen dropped).
• One stateKey per search. If you run several schedules on one account (say python and design), give each its own stateKey. Runs that share a key share the list, so one run would use up the other's new jobs. A new stateKey starts fresh and returns everything once.
• The run summary is in the OUTPUT record: { jobsSaved, newJobs, alreadySeen, stateKey, stateSize, warnings } (warnings lists every source or company board that failed or could not be understood; it is also there when onlyNew is off).
• Delete the record seen-<stateKey> in the remote-jobs-state store to reset a feed.

{ "keywords": ["python"], "onlyNew": true, "stateKey": "python-jobs" }

Company career boards

companyBoards adds the open roles of specific companies straight from their applicant tracking system:

Entry	Meaning
greenhouse:stripe, https://boards.greenhouse.io/stripe	Greenhouse board of Stripe
ashby:linear, https://jobs.ashbyhq.com/linear	Ashby board of Linear
lever:spotify, https://jobs.lever.co/spotify	Lever board of Spotify (EU-hosted jobs.eu.lever.co URLs work too)

The name is the last part of the company's careers URL. Records use source = greenhouse, ashby or lever, the original job url, and salary when the board publishes it. Company boards list all jobs, so remoteOnly (default on) keeps only roles advertised as remote (remote workplace type or "Remote" / "Anywhere" in the location); switch it off for every role. The keyword, location, date and salary filters apply to each board, and maxItemsPerSource caps the roles kept per board (so a large board with hundreds of on-site jobs still yields all of its remote ones). A misspelled name or a company that left the platform only logs a warning, the rest of the run continues. To pull from company boards only, empty the sources list (that stays "company boards only" even if every board entry was mistyped, so a typo never turns into a full, charged run over all the remote boards). At most 100 boards per run. Ashby and Lever do not return a company name, so it is derived from the name you gave ("acme-corp" becomes "Acme Corp").

{
  "sources": [],
  "companyBoards": ["greenhouse:stripe", "ashby:linear", "lever:spotify"],
  "keywords": ["engineer"],
  "onlyNew": true,
  "stateKey": "target-companies"
}

Example: fresh Python/data roles, deduplicated

{
  "keywords": ["python", "data engineer"],
  "postedWithinDays": 14,
  "maxItems": 300
}

Example: one board only, everything

{ "sources": ["remoteok"], "maxItems": 500 }

Example: well-paid, worldwide, lightweight feed

{
  "locations": ["Worldwide"],
  "minSalaryUsd": 120000,
  "includeDescription": false,
  "maxItems": 200
}

Output

One record per job (or per merged duplicate group). Example, taken from a real run:

{
  "id": "b8fbb55311035e5e880d7be66365a0b8e1aee54e",
  "source": "himalayas",
  "sourceJobId": "https://himalayas.app/companies/nvidia/jobs/senior-software-architect-networking-ai",
  "title": "Senior Software Architect, Networking AI",
  "company": "NVIDIA",
  "companyLogo": "https://cdn-images.himalayas.app/9jkqq85jdsq9vtt2fuhb5vq8y2zy",
  "companyUrl": "https://himalayas.app/companies/nvidia",
  "location": "United States",
  "isWorldwide": false,
  "employmentType": "Full Time",
  "seniority": "Senior",
  "category": "Developer",
  "tags": ["Software-Architecture", "AI", "Networking", "Data-Center-Architecture", "..."],
  "salaryMin": 184000,
  "salaryMax": 356500,
  "salaryCurrency": "USD",
  "salaryPeriod": "yearly",
  "salaryText": "$184,000 - $356,500 / yearly",
  "description": "NVIDIA has been transforming computer graphics, PC gaming, and accelerated computing for more than 25 years. ... We are seeking a Principal SW Architect, Networking to build the future of AI data-centers. ...",
  "url": "https://himalayas.app/companies/nvidia/jobs/senior-software-architect-networking-ai",
  "applyUrl": "https://himalayas.app/companies/nvidia/jobs/senior-software-architect-networking-ai",
  "postedAt": "2026-09-18T09:43:51.000Z",
  "scrapedAt": "2026-09-18T09:50:45.683Z",
  "alsoSeenOn": []
}

Field notes:

• isWorldwide — heuristic: true when the location text is empty or says something like "Remote", "Anywhere", "Worldwide" or "Flexible / Remote" (i.e. no specific country/region restriction).
• salaryText — always the best available human-readable salary string, even when a source only publishes free text (e.g. Remotive's "$90 - $150 /hour"); salaryMin/salaryMax/salaryCurrency/ salaryPeriod are the parsed-out numeric fields when they could be extracted.
• alsoSeenOn — when dedupe is on and the same job (matched by normalized company + title) appears on more than one board, this lists the other sources it was found on; the saved record is the richest copy (has salary data > longer description > most recently posted).
• description — omitted (null) when includeDescription is off; otherwise stripped to plain text or kept as raw HTML per descriptionFormat.

Pricing

Pay per event — you are charged only for what is saved:

Event	When
job-record	Each job saved to the dataset (after filters and deduplication)

Filtered-out jobs, merged duplicates and failed source requests are free. Set Max total charge on the run to cap spend; the Actor stops cleanly when the cap is reached. See PRICING.md for the exact per-event price and rationale.

Scheduling & integrations

Run on a schedule (e.g. daily or a few times a week — see "Limits & good citizenship" below for why not more often), then pull results via the Apify API, webhooks, Google Sheets, Make, n8n or Zapier. Turn on onlyNew so each scheduled run delivers only the jobs that are new since the previous one. Each record has scrapedAt, so scheduled runs can also be appended into a time series.

Limits & good citizenship

The Actor throttles itself to at most 2 requests/second per host, well under any of these boards' limits, retries on 429/5xx with back-off, and uses a descriptive User-Agent (remote-jobs-aggregator/1.0). A source that errors out (timeout, unexpected payload, rate limit) only loses its own jobs for that run — it logs a warning and the run continues with the other 7.

Attribution & terms

Company board data comes from the public job-board endpoints of Greenhouse, Ashby and Lever (no login), and the Actor is not affiliated with them either.

This Actor is not affiliated with RemoteOK, Remotive, Himalayas, Jobicy, Arbeitnow, We Work Remotely, Working Nomads or The Muse. Every record includes url (the original job page) and source; if you republish listings, link back to the original posting and credit the source board, and review each source's own API terms before high-volume or public-facing use. Two sources are explicit about this in their own API responses:

• RemoteOK: "Please link back (with follow, and without nofollow!) to the URL on Remote OK and mention Remote OK as a source, so we get traffic back from your site. If you do not we'll have to suspend API access."
• Remotive: "Please do not submit Remotive jobs to third Party websites, including but not limited to: Jooble, Neuvoo, Google Jobs, LinkedIn Jobs. Please link back to the URL found on Remotive AND mention Remotive as a source... there is absolutely no need to request Remotive Job data too frequently... we advise max. 4 times a day."

In short: attribute every source you display, don't resubmit Remotive's jobs to other job aggregators (Google Jobs, LinkedIn, Jooble, etc.), and don't schedule runs more than a few times a day. No personal data is collected beyond what each board already publishes in its public job ads (recruiter/company contact info that a board itself includes in a listing may appear in description).

Code example

A runnable Python script (only requests needed) that calls this Actor and writes a CSV: remote-jobs/python_jobs_feed.py — print a de-duplicated feed of remote jobs matching your keywords.

FAQ

Is this affiliated with any of the 8 job boards? No. It's an independent tool that reads each board's public API/feed.

Why do some jobs have salaryMin/salaryMax as null? Not every board publishes structured salary data. salaryText still shows whatever free-text salary info was available, if any.

Why did Remotive only return ~15 jobs? That's the board's own current free-API limit, not a bug in this Actor — verified live, and unaffected by search/category params.

A job appears twice in my results. Turn on dedupe (default on). Duplicates are matched by normalized company + title, so a very different title or company spelling on one board (e.g. a staffing agency repost) can occasionally slip past the matcher.

Something wrong or missing? Open an issue on the Actor's Issues tab — fixes usually ship within a day.