praxyjobs
Browse documentation
On this page

Documentation

The Praxy Jobs API

A jobs API built on one rule: every job comes straight from the employer's own ATS board and carries the receipt to prove it. 38 platforms (see them all), millions of live postings, and a source_slug + source_endpoint on every row so you can verify any listing at the source. No aggregator middlemen, no dead links resold as fresh.

Base URL

All endpoints live under /v1 on the public origin:

Example
https://praxyjobs.com/v1

Responses are JSON. Errors follow RFC 9457 problem+json — a status, title, and human-readable detail on every failure.

Try it in 10 seconds

Example
curl "https://praxyjobs.com/v1/jobs/search?q=data+engineer&country=DE&limit=3" \
  -H "X-API-Key: $PRAXY_JOBS_KEY"

Production requests use your subscription-backed API key. To evaluate the data before subscribing, use the capped ten-job preview on the homepage.

Authentication

Your API key is created after Dodo confirms an active subscription. Reveal it once from the dashboard, store it as a secret, and send it with every production request:

Example
curl "https://praxyjobs.com/v1/jobs/search?q=site+reliability+engineer&include_description=true&limit=1" \
  -H "X-API-Key: $PRAXY_JOBS_KEY"

What's in a job

Stable job_key, title, organization, direct apply URL, status with close reasons, first/last-seen timestamps, derived locations, employment types, salary when the source publishes it, a career-family taxonomy, and the verifiability block: source, source_slug, source_endpoint, directory_verified_at. Fields we don't know are omitted, never guessed — absence means unknown.

Where to next

  • Full API reference — every endpoint and filter, including the title boolean grammar and the active/expired/modified feeds.
  • Rate limits — the actual per-endpoint numbers, no fine print.
  • MCP server — the same index inside Claude or Cursor, keyless.
  • OpenAPI spec — machine-readable, straight from the running API.
Was this page helpful?
YesNot yet