Remote Jobs Feed API — Himalayas, Remotive, Arbeitnow, We Work Remotely & HN in one schema

You want a live feed of remote jobs, and the good sources are five different APIs. Himalayas paginates by cursor, Remotive returns everything in one shot, Arbeitnow pages 250 at a time, We Work Remotely is RSS per category, and Hacker News Who is hiring? is free text. This Actor queries all five in a single run and returns one deduplicated remote-jobs dataset with the same 22 fields on every row.

No API key. No proxy. No browser. No login.

Himalayas ──┐
Remotive  ──┤
Arbeitnow ──┼──►  dedup by company+title  ──►  one schema, alsoSeenOn trail
WWR       ──┤
HN hiring ──┘

What problem this solves

Running five scrapers and merging their output by hand is the usual way to build a remote job feed. The merge is the annoying part: the same role is often carried by two boards under slightly different titles, and every feed names its fields differently. This Actor does the merge, and records which other feeds carried the same job in alsoSeenOn, so you keep the evidence instead of silently throwing a duplicate away.

Use cases

• Run a remote job board or newsletter for a niche — Python, design, EU timezones — without maintaining five integrations that each break on their own schedule.
• Remote job alerts at scale — schedule a daily run with postedWithinDays: 1 and you get only what is new since yesterday.
• Labour-market research — who is hiring remotely, for what, at what salary, over time.
• Feed an LLM or RAG index with fresh, structured, permissively-sourced job text.
• Candidate sourcing in reverse — see which companies are posting remote roles this week and in which categories.

Sources

Himalayas

himalayas.app/jobs/api — public JSON with cursor pagination, 20 jobs per page. The best of the five for structured fields: it is the main contributor of salaryMin / salaryMax / currency, seniority and locationRestrictions.

Remotive

remotive.com/api/remote-jobs — the public API Remotive publishes for third parties. Returns its whole current list in one request, and that list is small (17 jobs on 2026-09-10). Jobs keep their original url; please credit remotive.com when you republish.

Arbeitnow

arbeitnow.com/api/job-board-api — public job-board API and by far the deepest feed here (250 per page). Carries some on-site European roles as well as remote ones; those come back with isRemote: false.

We Work Remotely

The public per-category RSS feeds. Note that category slugs are not guessable — remote-marketing-jobs and remote-sales-jobs return 403; the working slug is remote-sales-and-marketing-jobs.

Hacker News "Who is hiring?"

hn.algolia.com/api/v1, the official public HN search API, reading the current Ask HN: Who is hiring? thread. Off by default, because the thread is free text: titles are the comment's first line and company names are best-effort. Turn it on when you want the startup roles that never reach a job board.

Input

{
  "keywords": ["python", "data engineer"],
  "sources": ["remotive", "himalayas", "arbeitnow", "weworkremotely"],
  "locationKeywords": ["Europe", "Worldwide"],
  "postedWithinDays": 14,
  "excludeKeywords": ["senior manager"],
  "includeDescription": false,
  "deduplicate": true,
  "maxItems": 200,
  "maxPagesPerSource": 3,
  "requestDelaySeconds": 1
}

Leave keywords empty to pull everything the feeds currently publish.

Output — one row per remote job

{
  "source": "himalayas",
  "jobId": "https://himalayas.app/companies/netomi/jobs/software-development-engineer-in-test-i",
  "title": "Software Development Engineer In Test - I",
  "companyName": "Netomi",
  "companyLogo": "https://cdn-images.himalayas.app/...",
  "url": "https://himalayas.app/companies/netomi/jobs/...",
  "applyUrl": "https://himalayas.app/companies/netomi/jobs/...",
  "locationText": "India",
  "locationRestrictions": ["India"],
  "employmentType": "Full Time",
  "category": "Developer",
  "tags": ["SDET", "QA-Automation-Engineering"],
  "seniority": "Mid-level",
  "salaryText": null,
  "salaryMin": null,
  "salaryMax": null,
  "currency": null,
  "postedAt": "2026-09-09T14:12:17+00:00",
  "descriptionText": null,
  "isRemote": true,
  "alsoSeenOn": ["remotive"],
  "scrapedAt": "2026-09-10T10:31:02.884000+00:00"
}

alsoSeenOn lists the other feeds that carried the same job when deduplicate is on. RUN_SUMMARY in the key-value store records how many rows each feed returned and any feed that failed.

The dataset ships with three views: overview (including alsoSeenOn), salary (salaryMin, salaryMax, currency) and full.

Honest limitations

• Field coverage differs by feed. Salary comes mostly from Himalayas; employmentType is missing from We Work Remotely; Hacker News has no structured fields at all. Fields a feed does not publish are null, never guessed.
• Keyword filtering is client-side. The feeds are fetched first and filtered after, so a narrow keyword with a low maxPagesPerSource can return very little. Raise maxPagesPerSource before concluding there are no matches — a measured run with keyword python returned 8 rows from 692 jobs fetched.
• Deduplication is heuristic — normalised company + title. Two genuinely different roles with the same title at the same company collapse into one row; turn deduplicate off if that matters.
• Feed depth is finite and wildly uneven. In a measured run: Arbeitnow 250, Hacker News 100, We Work Remotely 38, Himalayas 20, Remotive 18. So maxPagesPerSource buys very different amounts of data per feed.
• This is a current openings tool, not a historical archive. The feeds publish what is open now. If you want history, run it on a schedule and keep your own dataset.
• Remote by nature. These are remote job boards; Arbeitnow is the only one carrying meaningful on-site volume.
• These feeds are free services. They can rate-limit or change shape. The Actor retries with backoff and records per-feed errors in RUN_SUMMARY rather than failing the whole run.

Pricing shape, and why

This Actor charges a small start fee plus a per-job fee. That is deliberate: its cost is driven by fetching 700–900 jobs, but a narrow keyword may push only 8 rows. Pure per-result pricing would make a precise, well-filtered search — the useful kind — the one that pays least.

Attribution

robots.txt was checked for every host; none of these paths are disallowed. There is no login, captcha solving, proxy rotation or browser automation anywhere in this Actor. When you republish jobs, keep the original url and credit the source board.

FAQ

What is the best API for remote jobs? There is no single one. Himalayas has the best structured fields, Arbeitnow has the most volume, Remotive has the best-known brand and the smallest list. This Actor's answer is to read all of them and hand you one schema.

How do I build a remote job board? Schedule this Actor daily with postedWithinDays: 1, deduplicate: true and your niche in keywords, then push the dataset into your site. alsoSeenOn tells you how widely a role is already listed, which is useful for ranking.

Does this include work-from-home jobs that are not tech? Partly. Arbeitnow and We Work Remotely carry non-engineering categories (support, marketing, sales, design); Himalayas and Remotive skew technical. Use keywords and excludeKeywords to shape it.

Is there a free remote jobs API? All five sources are free and public — that is why they were chosen. What you are paying for here is the querying, pagination, normalisation and deduplication across all five, not access to the data.

Can I get salary data? Where the board publishes it. Use the salary dataset view: salaryMin, salaryMax and currency are parsed out of the raw salaryText where possible. Himalayas is the main contributor; expect null on many rows.

Do I need a proxy? No. This Actor exposes no proxy input.

Changelog

See CHANGELOG.md.

Support

Found a bug, or need a field this Actor does not return yet? Open an issue on the Actor's Issues tab, or email pablodevigoalertas@gmail.com. Issues are read first.