How this is built

Rules I did not have to write down.

These are not values on a wall. Each one costs something, and that is how you can tell it is real.

Open licences, attribution attached

Every question comes from an openly licensed dataset - eleven of them - and carries its source and licence with it, all the way to the API response. It would be cheaper to strip that field. It stays, because a question without a source is a question you cannot defend.

"attribution": {
  "author": "uberspot/OpenTriviaQA contributors",
  "source": "opentriviaqa",
  "license": "CC-BY-SA-4.0",
  "licenseUrl": "https://creativecommons.org/licenses/by-sa/4.0/"
}

The real attribution block every API answer carries - copied from a live response, 11 August 2026.

A free tier with no card

500 calls a day on the full production API, no card, no trial clock counting down. If you cannot build something real before deciding to pay, the free tier is a demo, not a free tier.

Numbers rounded down, never up

The question counts in the app store listings are lower than what the apps actually hold. Rounding up is free and nobody would check. That is exactly why it is not worth doing.

Example: the base held 1,173,456 questions on 26 May 2026 - this site says "1.1M+".

One person, and it says so

No "our team", no stock photos of an office. A one-person studio has real limits and pretending otherwise only moves the disappointment later.

The moat is the craft, not the content

The questions come from open datasets, so anyone can have them. What takes years is the curation, the formats, the creator and the care. That is the part worth building.