The anti-hallucination layer, as an API.

Your AI product cites facts. Sometimes it invents them. POST a claim to our verification API and get back a verdict, real sources, and a confidence score, drawn from a corpus of claims that's already been argued over in public.

POST /v1/verify
{ "claim": "power posing raises testosterone" }

200 OK
{ "verdict": "REFUTED",
  "confidence": 38,
  "for": 1, "against": 2,
  "sources": [ ...real papers, with DOIs... ] }
Pricing

Per claim retrieved. Two prices, because there are two costs.

A claim someone has already argued over is a lookup. A claim nobody has checked yet runs the whole engine: query generation, retrieval from the open literature, and the judgement pass. Charging one blended rate for both would mean overcharging everyone who hits the corpus and undercharging everyone who grows it.

What you asked forPer claimPer 1,000
Retrieved from the corpusAlready checked. Semantic match against the existing record, returned with its receipt. $0.0001$0.10
Constructed by the engineNobody has checked it. Queries generated, papers retrieved and filtered, evidence weighed, receipt minted and added to the corpus. $0.005$5.00
Deep searchThe same engine across every database we hold: Europe PMC, OpenAlex, Semantic Scholar, Crossref, PubMed and arXiv, triangulated and deduplicated. More citations; not always a different answer. $0.009$9.00
Cheapest Full Fact Checks
in the Industry
It gets cheaper the longer you use it
Every claim the engine constructs becomes a retrieved claim for whoever asks next, so the corpus hit rate climbs as you use it. The same volume of checks costs less next year than it does this one.
The field

What everyone else charges to answer the same question.

There is no other API that takes a claim and returns a verdict with real papers behind it, so this compares the things people actually wire up instead: a model grounded in web search, a hallucination evaluator, a retrieval API, and the one free fact-check index. Prices are per request, converted to the same unit as ours.

Feature trust me. Gemini +
Search		Perplexity
Sonar		Patronus
AI		 ExaGoogle Fact
Check Tools
Answers a claim nobody has checked before ✓ ✓✓ ✓✓ ×
Returns a verdict, not just links or a score ✓ ×× ×× ✓
Cites peer-reviewed papers, with DOIs ✓ ×× ×× ×
Weighs the evidence against the claim too ✓ ×× ×× ×
Verifies quotes word for word ✓ ×× ×× ×
Drops retracted sources ✓ ×× ×× ×
Answer arrives as a public page anyone can challenge ✓ ×× ×× ✓
Gets cheaper as the corpus grows ✓ ×× ×× ×
Free to try 1,000 claimson us, no card 5,000/monone $10 credit$20 + $10/mo free tier
Per claim $0.0001 corpus$0.002 new $0.014+ tokens $0.005–0.014+ tokens $0.010–0.020per eval $0.007search only $0if it exists

Each provider's own published price, checked 30 July 2026, converted to one request. Gemini grounding $14 per 1,000 searches, Perplexity Sonar $5 to $14, Patronus $10 to $20 per 1,000 evaluations, Exa $7 per 1,000 searches.

Google's Fact Check Tools API is free but only looks up claims a human fact-checker has already published: in our own testing about 84% of tested claims came back empty.

Factiverse and scite are the nearest peers and neither publishes a per-call price. We are not affiliated with any of them. Spot a number that has moved? Challenge it and we will fix it in public.

Getting a key

Sign up, get 1,000 claims on us* and load a wallet.

  1. Create an account. Email and password, same account as the app. We need a real identity behind a key so a leaked one can be revoked and so abuse has a name attached to it.
  2. Generate a key. One click in your dashboard. We show it once and store a hash, so if you lose it we cannot recover it and neither can anyone who reads our database. Send it as Authorization: Bearer tmb_live_…
  3. Spend your free credit. Every new account gets $0.10, no card. *That is 1,000 claims already in the corpus, or 20 the engine has to construct from scratch, or around 90 on a realistic mix: enough to wire it up, run a real workload through it, and decide.
  4. Load the wallet. When the credit runs out, top up between $5.00 and $50.00 at a time. No subscription, no seat fees, no minimum monthly spend. The balance does not expire.
  5. Watch it draw down. Every response carries X-TMB-Cost and X-TMB-Balance, so you can meter your own users without polling us. At zero we return 402 with the top-up link rather than silently degrading the answer.
Why prepaid rather than a monthly bill
Because a claim costs fractions of a cent, a monthly invoice would mostly be an invoice for the invoice. Prepaid also means you can never be surprised: the ceiling on what we can charge you is the money already in the wallet, which is a harder cap than any rate limit. There is a rate limit as well, and we publish it: 60 requests a minute and 1,000 an hour, per account rather than per key. That is roughly fifty times what our heaviest modelled customer needs, and if you need more, write to us and say so.
The honest part

What the API does not include.

The API is the engine's judgement, priced per claim. It is not the product our subscribers pay for, and you should know the difference before you wire it up.

No receipt page. You get JSON: the verdict, the sources for and against, a confidence score and a summary. You do not get a public trustmebro.to page for the claim, and no share link comes back. The receipt, and the argument that happens on it, stay with the app.
No Deep Search. Every API call runs the standard path. The wider cross-provider sweep, the analytical lens, and the 127 schools of thought are Pro features in the workspace, not parameters on this endpoint.
No draft review. Whole-document claim detection, the argument critique with its counter-case, and the logic grade are not exposed. One claim per call, up to 600 characters.
No formatted citation string. You get the bibliographic record for every source: authors, year, title, venue, DOI, URL, citation count. Building that into APA, MLA, BibTeX or RIS is one function in your own code, and the workspace is where we do it for a person instead.
No interface. No extension, no side panel, no library, no export. If you want a person to use this rather than a program, they want the app.
Which is why a check through the API costs a fraction of a check on a plan: it is the same engine sold wholesale, without the surfaces that make it usable by a human. If you are checking claims inside your own product, that is the right trade. If you are the one reading the answer, the plans are.
Join the waitlist
The API opens after launch. Leave an email and you're first in line. Already have an account? Open your wallet and generate a key.
Or join the general early-access waitlist and get the API, the extension and the full wiki as each one opens.