Skip to content
TruePointTruePointData

Documentation

Quickstart

What the API does, what a call costs, and how to make your first one.

The TruePoint data API answers two questions: who is this company, and who works there. You send a partial identifier — a domain, an email, a name plus an employer — and get back a structured record where every field carries its own provenance.

It is priced in credits and metered on results. A call that returns nothing costs nothing, which means our coverage gaps show up in our match rate rather than on your invoice.

Three things worth knowing before you integrate

  • Company matching is free. Resolve domains to ids as often as you like; it is rate-limited, not billed.
  • Every response carries field_provenance. A value with three corroborating sources seen last week is not the same asset as a single-source value from last year, and the API refuses to pretend otherwise.
  • Suppressed people are absent, not redacted. Someone who has opted out comes back as a no-match, on every endpoint, with no flag to override it.

Your first call

Start with company matching, because it is free and it tells you immediately whether your identifiers resolve to the entities you expect.

bash
curl "https://api.truepoint.in/api/v1/public/company/match?domain=acme.com" \
  -H "Authorization: Bearer $TRUEPOINT_API_KEY"

A match returns a stable company id. Store it: it is the cheapest way to re-query later, and it survives a company changing its name or domain.

Then enrich

bash
curl -X POST https://api.truepoint.in/api/v1/public/company/enrich \
  -H "Authorization: Bearer $TRUEPOINT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"id":"cmp_2b81"}'

Every billed response includes credits_charged, so your own metering can reconcile against ours per call rather than per invoice. If those two numbers ever disagree, that is a bug on our side and we want to hear about it.

Where to go next

  • Authentication — how keys work, and what to do when one leaks.
  • Errors — the full problem+json vocabulary, including which failures are billed (none of them).
  • Pagination and rate limits — how to page a search without paying for results you discard.
  • Confidence and provenance — how to read the field_provenance block, and what each source class means.