Documentation

Checking many at once

Up to 500 in one call, in the order you sent them.

Checking fifty graduates should not be fifty requests and fifty scraped web pages.

POST /api/v1/verify/batch
Authorization: Bearer <your API key>
Content-Type: application/json

{ "codes": ["ACME-MERN-NOV26-0007", "ACME-MERN-NOV26-0008"] }

One row per query, in the order you sent them, so the answer lines up against your spreadsheet. Duplicates are kept rather than collapsed, for the same reason.

{
  "rows": [
    {
      "query": "ACME-MERN-NOV26-0007",
      "found": true,
      "status": "published",
      "valid": true,
      "recipientName": "A Sharma",
      "issuer": { "name": "Acme Institute" }
    },
    { "query": "ACME-MERN-NOV26-0008", "found": false }
  ],
  "summary": { "requested": 2, "found": 1, "valid": 1, "notFound": 1 },
  "tookMs": 180
}

tookMs is the measured server-side lookup. The published budget is 2 seconds for 500 codes, and returning the measurement means you can check that rather than take our word for it.

Without an account

POST /api/public/credentials/verify-batch

Up to 25, no key, no charge. The split is about rate limits rather than generosity: an HR team pasting a shortlist is ordinary use; thousands of checks a day needs its own pool, which is what a key is for.

What it will not do

It reveals nothing a single check would not. Unpublished credentials answer as not found, and a credential whose issuer chose a private check link answers only to that link's token — otherwise the printed IDs could be walked five hundred at a time.

Looking people up by name and email is scoped to your own organisation. Answering that across every issuer would make this a people-search engine: post an email, receive every credential that person holds anywhere. A credential is public by its ID, which the recipient was given — not by the recipient's identity.