IRS Form 990 data, normalized and served as infrastructure.
Nonprofiteer turns the public Form 990 corpus into a clean, queryable dataset with a changed-since sync feed on top — an affordable, open alternative to the incumbent directories. It models the filings and nothing else: what you build on them is yours.
You cannot own public records. The charge is for access and infrastructure, never for the facts.
The corpus
- Organizations
- 1.97M
- Exempt organizations with a BMF record
- People
- 10.7M
- Officers, directors, trustees, key employees
- Filing years
- 2022–25
- Form 990 returns parsed to date
Measured August 2026. Rebuilt monthly from the IRS Exempt Organization Business Master File and the Form 990 e-file corpus.
What is in the data
Available now
- Organization identity: EIN, names, exemption and subsection codes, group-exemption parents
- Normalized addresses, linked to the organization and to the filing they came from
- Form 990 Part VII people, with the filing each name was read from
- Part I top-line money: total revenue and expenses, contributions received, grants paid
- Provenance on every record, back to the source document
Not yet
- Itemized financials — the Part VIII/IX/X line items and Schedule I grant lists
- Form 990-EZ and 990-PF returns; full Form 990 only for now
- Anything not published by the IRS: no scraped, purchased or contributed data
The API
A JSON:API feed at /api/v1/sync, keyset-paginated
and ordered so a consumer can resume exactly where it stopped. Every record carries whether
it was created, superseded by an amendment, or tombstoned, so a mirror stays correct
without re-walking the corpus.
Requests are authenticated with a per-consumer API key and metered against published limits, which the reference documents. Access is currently limited to the project's first integration while the pipeline is in early release; there is no open signup yet.