How it works

How we actually know this isn't just another AI guessing

Fair question, and worth answering properly instead of asking you to just trust us. Here's exactly what happens between a claim and a verdict, written so you can explain it to someone else, not just take our word for it.

Think of it like a research assistant who never works from memory

When you ask an AI a question directly, it answers from memory — patterns it picked up during training. That's fast, but it means the AI is reciting something it half-remembers, and half-remembered things are exactly where AI confidently states things that aren't true. That's what "hallucination" means: not lying, just recalling wrong with full confidence.

For every fact in this dataset, we don't ask an AI what it remembers. We send it to go find the real answer, live, from the actual source, the way you'd send a research assistant to go pull the real document instead of answering from the top of their head.

01

A claim exists

Someone states a fact — a founding year, a court ruling, a lab result, a company's reported revenue.

02

A live lookup, not a memory

We send Claude to the real source right then: a government database, an actual SEC filing, the real text of a court opinion. Nothing is answered from recall.

03

Read, not recalled

Its job is to find the specific fact on that real document and write down exactly what it says, not to answer from what it thinks it knows.

04

Verdict, plus the receipt

The claim is marked match, diverges, or unverifiable, and the exact source link ships with the row so anyone can check it themselves in under a minute.

Here's a real one

Row MV02 in the Mathematical Claims Verification dataset. The claim: that 2³¹−1 (2,147,483,647) is prime, first proved by Leonhard Euler in 1772. Here's what the live lookup actually found, and, right below it, what happened when we separately asked Claude the identical question completely cold, no source, no lookup, just memory.

Row MV02 · math-verification live source
claim2³¹−1 is prime, proved by Euler, 1772
verification_source_urlwikidata.org/wiki/Q773522
what the live source saysP61 discoverer = Leonhard Euler · P575 date = 1772
agreementmatch
Same question, asked cold, no source memory only
questionWho first proved 2³¹−1 is prime, and when?
Claude's answer"...proven prime by Édouard Lucas in 1876..."
verdicthallucinated

Same underlying fact, same model. Looked up live against a real source: correct, with a link you can click. Answered from memory alone: a confidently wrong name and a confidently wrong year. That gap is the entire reason this business exists.

Where we're honest about the risk

There's still a step in the middle done by AI: something has to actually find the right figure on a real document and write down what it says. Documents are messy, so there's a small, real chance of grabbing the wrong line or misreading a table. We don't pretend that risk is zero.

That's exactly why every row ships the source link instead of asking for your trust, why several products deliberately check two independent sources instead of one so a mistake in one gets caught by the other, and why a claim that can't be confirmed is marked unverifiable rather than forced into a guess. Verify us. Don't just trust us.

What happens when you skip the lookup

We tested this directly: every claim in the catalog, asked to Claude completely cold, no source, no tools, exactly like a normal chat. The error rate depends heavily on the domain.

Hallucination rate when asked from memory alone

by product, real test results
Citation-Graph Verification
48%
Clinical Trial Outcomes
18%
FDA Drug/Device Safety
14%
SEC EDGAR Financials
7%
Historical Event/Date Facts
0%

Across all 703 claims tested this way, the overall rate was 10.1% hallucinated, 73.8% correct even from memory, and 15.4% honestly hedged rather than guessing. Full numbers are in every product's README.

Want the fuller technical writeup?

Sourcing standards, the null-over-guess rule, and exactly which primary sources back each product.

Read the methodology