Non-dilutive funding, and a directory that is true this week.
Grants and fellowships with the parts nobody else records: the real eligibility rules, how each program actually pays out, its policy on AI-written applications, and the date every fact was last checked. Where nothing has been verified, it says unknown — a confident wrong answer is what makes a directory dangerous.
Free, complete, and no account required. 39 programs, each linking straight to the funder's own form.
What a list gets wrong
This started from a public directory of 41 microgrants. It was wrong about at least five of them — programs listed as rolling that had closed months earlier, a funder that had renamed and moved, an amount quoted in the wrong currency. A directory is a list. What an applicant needs is a list that is true this week, and nobody was maintaining one.
Records that were simply wrong
Closed cycles listed as rolling, one three months past a deadline. A program split into two entries that were one. An official application link that pointed at a different program's form.
Closed on eligibility alone
Ruled out for one 40-year-old non-student outside North America, on age, country or student status — and finding that out took hours of reading pages one at a time. The filter that mattered existed nowhere as structured data.
The payout rail nobody records
An applicant nearly formed a US LLC to receive a $924 grant — a permanent Form 5472 filing obligation carrying $25,000-a-year penalty exposure. The right answer was a Payoneer account. No other directory has a payout-rails field.
One project, aimed six ways
The same technical work had to lead with robot perception for one reviewer, with consent and evidence for another, with deliberately-not-a-business for a third. Getting that wrong is how a strong project loses to a weaker one.
What actually decides whether you win
Not better prose. Four things do most of the work, in this order — and the biggest of them is the free part.
-
Not wasting applications on programs that were never possible · free here
Most of the value, and it costs nothing. Filter by your own age, country and student status; every program that drops out names the exact rule that removed it.
-
Applying at the right moment
A cycle missed by three months is a year lost. Every record shows when it was last checked and says so loudly once it is overdue.
-
Framing to the funder
The same project, correctly aimed, is a different application. What a funder actually wants is drawn from what they have said in public, not from a guess.
-
Asking for the right amount
Under-asking loses money on funded applications — and can lose the application itself by signalling low ambition.
Prose quality is fifth, and it is the thing every other tool sells.
Start with the directory
It is deliberately complete and deliberately terminal: it tells you where to apply, and then it sends you to the funder's own form. On its own it is already better than every directory that exists.
- Structured eligibility filtering, with the disqualifying rule named for everything it excludes.
- Payout rails, AI policy and amount ranges as real fields — or a visible unknown.
- A last-verified date on every program, degrading loudly rather than going quietly stale.