The full upstream README, mirrored here for reference. Install config, tool schemas, adoption signals, and an original overview live on the Federal Grants — find funding you are actually eligible for listing page.
Runs entirely in your browser. No account, no backend, nothing uploaded or stored.
Top 5 matches free. Unlock the full ranked list for $2 in USDC — paid directly to the operator's wallet and verified on-chain in your browser. No account, no card, no processor.
Maintainer: payment setup and wallet notes are in CRYPTO.md; the card-payment alternative is in ACTIVATE.md.
Stop reading grant announcements you were never allowed to apply for.
Every other Grants.gov tool hands you a list. This one hands you a shortlist — live federal opportunities ranked by whether your specific organization can realistically win them, with the reasoning shown for every single score.
Point it at your org profile. Get back grants you are actually eligible for, that you actually have time to write, sized to what you actually need, with a read on who has been winning the money before you.
A scraper returns rows. This returns judgment, and shows its work.
| Typical Grants.gov scraper | Federal Grant Fit Scanner | |
|---|---|---|
| Eligibility | Dumps the raw text field | Hard mechanical gate on the federal applicant-type codes |
| Deadlines | Prints a date | Days remaining vs. time you need to write it |
| Award size | Prints a number | Compared against what you asked for |
| Competition | Nothing | Who actually won this program before, and for how much |
| Scoring | None, or a black box | Every point traced to a named fact |
Grants.gov publishes machine-readable applicant-type codes on every opportunity. A 501(c)(3)
is code 12. A small business is 23. A federally recognized tribal government is 07.
This actor filters on those codes server-side, so you never even download the grants your organization is legally barred from winning. In live testing, a small business and a tribal government scanning the same database got result sets with zero overlap. That is the difference between a list and a shortlist.
It never guesses. If an opportunity publishes no applicant types, or only says
"Others — see text", you get NEEDS_REVIEW, not a false green light. Absence of evidence is
never reported as eligibility.
For each opportunity's CFDA program number, the actor pulls real reported federal awards from USAspending.gov and tells you the median award, the largest, and who received them.
That single number reframes an opportunity. A $250,000 request against a program whose median prior award is $47.7M is not a good fit — it means you would be competing against major research institutions. The scanner says so, in those words, instead of quietly scoring it as a match.
Every result is a ranked, CRM-ready record:
whyThisScore is the contract: no number appears without the fact that produced it.
STRONG_FIT (70+) — eligible, time to apply, well matchedPOSSIBLE_FIT (45–69) — worth a look, some frictionREVIEW_ELIGIBILITY — you may qualify, but only prose can confirm itWEAK_FIT — eligible but poorly aligned| Field | Description |
|---|---|
organizationType | Required. Nonprofit, small business, university, tribal, city/county/state, school district, housing authority, individual |
focusKeywords | Your mission terms, e.g. ["housing", "youth"] |
requestedAmount | What you need in USD — unlocks award-size fit and prior-winner comparison |
canCostShare | If false, matching-fund requirements are penalized heavily |
minDaysToApply | Your realistic prep time (default 14) |
includeUnrestricted | Also return grants open to any entity (default true) |
includeIneligible | Audit what got filtered out (default false) |
Both are official U.S. government APIs serving public data. No credentials, no scraping, no terms-of-service risk, and nothing to break when a site redesigns.
NEEDS_REVIEW means exactly that — the announcement text decides. Read it.null rather than a fabricated benchmark.This tool narrows the field. It does not write the proposal or guarantee an award.