Everything people actually ask, in one place. 20 answers, including the ones that are not flattering.
A citation points at a paper: authors, year, DOI, formatted for your essay. A receipt settles a question: the claim, the verdict, the evidence both ways and the confidence score, on a public link you can drop into any thread.
Two people. Someone mid-argument who knows they read it somewhere and needs the source in ten seconds, and someone mid-essay who has written forty confident sentences and needs the ones that will not survive a marker. Same engine, two surfaces: the popup for the first, the workspace for the second.
Any claim or question, typed on the site or highlighted on any webpage via the extension. Essay drafts by paste today; claim detection inside file uploads and Google Docs is on the roadmap.
Yes. The Trust API is the same engine behind one endpoint: POST a claim, get back a verdict, real sources and a confidence score. It is priced per claim rather than per seat, and it is what you want if you are building something that cites facts and cannot afford to invent them. Full pricing and access are on the API page.
Five checks a month are free, receipts and share links included, no card. Bro is $9.99 a month for 400 checks and 5 draft reviews; Pro is $29.99 and adds the side-panel workspace. Annual is twelve months for the price of ten.
No, and the product is built so we could not get away with it if we tried. Every citation is retrieved before it is judged, quotes appear only when they exist verbatim in the source text, retracted papers are caught by a nightly scan, and anyone can challenge any receipt with a counter-source.
We run a labeled evaluation before every release and publish all four numbers, including the unflattering ones: false-attribution rate, recall, verdict accuracy and quote fidelity. They live on the accuracy page with the dataset size and the run date, so you can tell whether a figure is current.
We say CONTESTED and show you both sides. A famous study that failed replication gets called exactly that. Flattening a live dispute into a clean answer is how you lose arguments to people who read the follow-up studies.
It comes back COMMON KNOWLEDGE with no sources attached. Bolting a paper onto 'grass is green' would be noise dressed up as rigour, so the engine refuses to do it.
You get INSUFFICIENT EVIDENCE and a plain sentence saying so. We would rather hand you nothing than hand you a citation that does not hold, and the receipt records the miss publicly rather than hiding it.
Yes, on any receipt. Bring a counter-source and the engine re-judges with your paper in evidence. Papers move verdicts; opinions do not.
Bro takes 2,000 words per upload and 10,000 words a month. Pro takes 5,000 per upload, batches up to 20 uploads into one run, and covers 100,000 words a month. Either way we cite about 10 claims per 1,000 words, so a 2,000-word essay comes back with roughly 20 citations attached.
One claim, one question, one citation, one receipt. Essay reviews meter separately, so a heavy essay week never eats your checks.
Essay text is kept for 30 days so you can reopen the review, then deleted. The claims and their verdicts stay, because that is what receipts are made of, but they are stored as claims rather than as your document.
A claim you send is checked, and a claim nobody has checked before becomes a receipt in the public corpus, exactly as it would from the extension. That is the trade that keeps the retrieved price near cost: your constructed claim is someone else's cheap lookup tomorrow. The receipt carries the claim and the evidence, never your key, your account or your end users. Do not send us personal data or anything confidential inside a claim, because a receipt is a public page.
No. A receipt carries the claim, the verdict and the evidence. It carries no identity, shared or not. Sharing publishes your connection to the claim, not your data.
It unlocks the side-panel workspace: whole-draft review, the argument critique with the counter-case, the analytical lens, a searchable claim library, and a deeper search on every check.
Any time, from your Stripe receipt email. No retention maze, no phone call, no three-step confirmation flow.
No subscription. You load a prepaid wallet and spend it a claim at a time: $0.0001 for a claim already in the corpus, $0.005 for one nobody has checked yet, and $0.009 if you ask for a deep search across every database we hold. That is $5 per thousand fresh claims, or $9 per thousand deep ones, and a corpus hit is a lookup rather than an engine run. Every new account gets $0.10 of credit with no card, top-ups run $5 to $50, and the balance does not expire. The ceiling on what we can charge you is the money already in the wallet.
Create an account, generate a key in your wallet, and send it as a bearer token. We show the key once and store only a hash, so a leak of our database yields no working keys and a key you lose cannot be recovered by anyone, including us. Keys are revocable, and five active keys is the limit. The API opens after launch; the waitlist on the API page is the queue.
We're cooking the most powerful verification tools on the internet, and a social space where debates get settled with receipts instead of volume. Join us for the ride: leave an email and you get early access to every new feature as it lands.