Jobvite Jobs API

Jobvite is a recruiting platform used by mid-size and enterprise employers, and its boards are the plain-HTML end of this pool: jobs.jobvite.com/<slug>, no JSON API, no feed. There are 2,043 Jobvite postings in this dataset — a number that moves hourly, so read it from /v1/meta rather than trusting this sentence.

Try it#

curl -G https://api.hyperjobs.io/v1/jobs \
  -H "Authorization: Bearer $HYPERJOBS_KEY" \
  --data-urlencode "source=jobvite" \
  --data-urlencode "limit=2"

{
  "data": [
    {
      "id": "jobvite:altura_oXY3bfwT",
      "title": "Staff Systems Engineer",
      "company": {
        "slug": "altura-systems",
        "name": "Altura Systems",
        "domain": "alturasystems.com",
        "website": "https://alturasystems.com",
        "logo": "https://media.licdn.com/dms/image/…",
        "industry": "Software Development",
        "employee_count": 5800
      },
      "location": {
        "remote": false,
        "countries": ["United States"],
        "cities": ["San Jose"],
        "regions": ["California"],
        "raw": ["San Jose, CA"]
      },
      "employment_type": null,
      "work_arrangement": null,
      "seniority": null,
      "salary": null,
      "skills": ["kubernetes", "linux"],
      "taxonomies": ["Software"],
      "apply_url": "https://jobs.jobvite.com/altura/job/oXY3bfwT/apply",
      "url": "https://jobs.jobvite.com/altura/job/oXY3bfwT",
      "source": "jobvite",
      "posted_at": "2026-07-13T11:20:44.000Z",
      "expires_at": null,
      "updated_at": "2026-07-14T11:20:44.000Z"
    }
  ],
  "total": 2,
  "limit": 2,
  "offset": 0
}

Note that apply_url and url differ — Jobvite hangs the application form off the posting path, which not every platform does. salary and employment_type are null because the board's HTML carries neither; more on the dates below.

How Jobvite boards actually work#

Two GETs, no auth. The important part is which two:

# 1. List — /search, and ONLY /search. `p` is 0-INDEXED, 50 per page
curl "https://jobs.jobvite.com/<slug>/search?p=0"
 
# 2. Detail — one per posting; the body only exists here
curl "https://jobs.jobvite.com/<slug>/job/<id>"

p is 0-indexed, so a loop starting at 1 quietly drops the first fifty jobs. page, pg, start and offset are all silently ignored and return page one.

It is not a JSON-LD board#

This is worth stating plainly, because Jobvite looks like it should be one and an early note in our own coverage docs said it was.

It isn't. Two tenants checked in detail — including a 200-posting board — carry zero ld+json blocks, and a generic JobPosting extractor returns null on them. Structured data exists on maybe a third of tenants. So HTML extraction is mandatory here, and we don't use the JSON-LD path at all rather than run two code paths that disagree with each other about the same board.

The practical consequence is on dates. The posting date exists only in the structured data, on the minority of tenants that ship it — the HTML carries none. Which brings us to:

What bites you#

A bogus id returns 200 and the board shell. So does an id from another tenant. There's no 404 to catch, and an unknown tenant is worse: it returns 200 and Jobvite's own support page. The only reliable "this is really a job" signal is the description container itself — if it isn't there, you don't have a posting, whatever the status code says.

jv-job-detail-meta is variable-length. It's a run of segments joined by a separator span, and tenants ship different runs: Dept | Loc | Req. Num. on one, Dept | Loc on another, Loc alone on a third. Parse it positionally and the third tenant's location becomes its department, on every posting, at 200. Take the requisition number by its own label and the last remaining segment as the location — that's the one segment every tenant has.

<h1> is the site name. The job title is <h2 class="jv-header">. Reach for the <h1> out of habit and every posting on the board is named after the company.

The description container nests. A lazy match to the first </div> truncates the body, so the div walk has to be balanced.

Or use ours#

Jobvite is small, and small platforms are where the effort-to-reward ratio is worst: the same class of problem as a big platform, spread across a fraction of the jobs.

You get the board enumerated through /search rather than the root that shows you half of it, ids taken from a 0-indexed loop, descriptions from a balanced walk, the location read by label rather than by position so a one-segment tenant doesn't report its city as a department, the shell-page 200s rejected instead of stored as empty jobs, and the company join.

And you get it deduplicated against the same role on LinkedIn or on another board — which on a source this size is a real share of the rows.

Refresh cadence varies by source; see coverage, and don't build anything that assumes uniform latency.

Query it with ?source=jobvite. Quickstart · Pricing