Remote Jobs Feed

Pull remote job listings from RemoteOK, Himalayas and We Work Remotely in one run, and get back a single clean feed: one schema, salaries you can actually compare, and the same job cross-posted to several boards merged into one record.

Most job scrapers hand you one board's raw shape. If you want more than one board, you get incompatible field names, salaries in different units, and the same job three times. This does that reconciliation for you.

Quick start

Click Run with defaults. You get 100 recent remote jobs from all three boards, deduplicated, with no configuration.

To narrow it down:

{
  "searchTerms": ["python", "backend"],
  "minSalaryUsd": 120000,
  "worldwideOnly": true,
  "maxResults": 50
}

What makes the output different

One schema across three boards. RemoteOK calls it position, Himalayas calls it title, We Work Remotely packs "Company: Role" into an RSS title. All three come back as title and company.

Salaries you can compare. Himalayas quotes some roles hourly and some annually, in varying currencies. RemoteOK uses 0 to mean "not disclosed". Everything is annualized into salary_min_usd_year / salary_max_usd_year.

Where a salary is quoted in a currency other than USD it is not converted, because a stale exchange rate silently corrupts data. The original figures are kept in salary_min_raw / salary_currency and salary_is_normalized tells you whether a number is comparable. You always know what you are looking at.

Cross-posted duplicates merged. The same job on several boards becomes one record. The most complete copy wins, the others are listed in also_on, and gaps are filled across copies — so if one board has the salary and another has the description, you get both. A typical run collapses 6-10 cross-postings.

A worldwide flag that works. "Anywhere in the World", "Worldwide" and "Global" are all recognized, across three boards' differing conventions.

Output fields

Field	Description
id	Stable id, prefixed with the source
source	remoteok, himalayas or weworkremotely
also_on	Other boards this job also appeared on
title, company, company_logo	Core identity
description_html, description_text	Full description, raw and flattened to text
categories, tags, seniority	Classification, unioned across boards when merged
employment_type	full_time, part_time, contract or internship
location_raw, location_restrictions, timezone_restrictions	Geography as each board states it
is_worldwide	True when there is no geographic restriction
salary_min_usd_year, salary_max_usd_year	Annualized USD, null when not comparable
salary_is_normalized	Whether the USD figures above are trustworthy
salary_min_raw, salary_max_raw, salary_currency, salary_period	The source's original figures
posted_at, expires_at	ISO-8601 UTC
apply_url, source_url	Where to apply

Input options

Option	Default	What it does
sources	all three	Which boards to pull from
maxResults	100	Jobs returned after dedupe and filtering; 0 for no limit
searchTerms	none	Keywords across title, company, tags, categories, description
matchAllTerms	false	Require every term rather than any
minSalaryUsd	none	Annual floor; excludes undisclosed and unconvertible salaries
maxAgeDays	none	Recency window; undated jobs are kept
employmentTypes	all	Filter to specific types
worldwideOnly	false	Only roles with no geographic restriction
deduplicate	true	Merge cross-posted duplicates
includeDescription	true	Turn off for a much smaller dataset

Typical uses

• Building a remote job board or newsletter without maintaining three separate integrations
• Salary benchmarking across remote roles, using figures that are actually comparable
• Feeding job descriptions to an LLM — description_text is pre-flattened with paragraph structure intact
• Recruiting research on who is hiring remotely, in what categories, at what pay

Notes

All three sources are public, documented endpoints rather than scraped HTML, so this reads published feeds instead of parsing markup. That means it does not break when a site adjusts its page layout.

If one board is unavailable, the run continues with the others and the status message names what failed. You are never charged for records you did not receive.

Pricing

$4.00 per 1,000 results. Platform usage costs are included — you are not billed separately for compute.