What a payer published, and on what date.

Each of these is a question an analyst already asks by hand, against files that may not be there next month. The right column names the call that answers it once the world is installed in your agent.

1. 01

   What did this payer publish for this code?

   The median, the quartiles and the row count behind them, for a billing code in a market, from the negotiated-rate file as it was published. Where the evidence sits below a reporting floor, or is absent, the answer says so instead of returning a number.

   CALL rate_position

2. 02

   How does that sit against the others?

   The same code across the payers in the corpus, and against the fee-schedule anchor, so a rate has something to be high or low against.

   CALL rate_comparison

3. 03

   What is the schedule underneath it?

   The Medicare Physician Fee Schedule and Clinical Laboratory Fee Schedule figure, with status indicator and the professional and technical split, so a global rate is not compared with a component.

   CALL fee_components

4. 04

   What did the hospital file?

   Price-transparency filings under 45 CFR 180, kept in their own tools and never mixed into payer rates, because a filed charge is not a published rate.

   CALL filed_charge, hospital_variance

5. 05

   Who is allowed to furnish and interpret it?

   The coverage article language on supervision and interpretation, and the procedure-to-procedure edits that decide whether two codes may be billed together at all.

   CALL interpretation_requirement, stack_check

6. 06

   Will the sentence survive review?

   Put the sentence you intend to write to the checker before it goes in the deck. It returns the figure the sentence rests on, or refuses it and names what would settle it.

   CALL claim_check, judgement_check

A published rate is not an allowed amount, not a paid amount, and not coverage. Ask this world whether a payer covers a code and it refuses by name, and says what evidence would settle it. That refusal is the feature: a sentence stopped before the deck is worth more than a footnote after the review. There are 25 refusal classes, each published in advance, so a caller knows what is off the table before asking.

The boundary between a payer rate and a hospital filed charge is held in the corpus itself rather than in a caveat, and CMS has issued a technical clarification on how these files are to be read.

Since 1 January 2026, CMS-0057-F has required impacted payers to decide prior authorisations inside 72 hours or seven days and to keep an audit-ready record of what was decided and why. We do not make those decisions. We hold the evidence one rests on, at the vintage it had on the date of service, so which version applied, and on what date still has an answer months later. The rule.

On the policy side the same problem is measurable. CMS rebuilds its coverage database in place with no changelog, so the only way to know what moved is to have kept both sides. Between two dated extracts, 22 policies were revised. The comparison is published on the reader. A competitor starting today cannot reconstruct it from today’s CMS endpoints. Read the archive.

Said plainly, because it matters more than a bigger number would: collection began in July 2026. That is about two months of vintages, not years. The mechanism is proven and the archive is young, and every month it continues is a month a later entrant cannot reconstruct.

What is a payer-published rate?

It is the negotiated rate a payer discloses in its machine-readable transparency file for a billing code and a contracted provider. It is what the payer published it pays when it pays. It is not the allowed amount, not what was paid on a specific claim, and not a statement that the payer covers the code.

Does a published rate tell me what will actually be paid?

No, and any tool that says otherwise is overselling the file. The published rate is one input into an allowed amount. Modifiers, place of service, plan design, member liability, bundling and downstream adjudication all sit between it and a remittance. What the published rate does establish is what a payer put on the record, and on what date, which is the part a reviewer can check months later.

Where does the rate data come from?

Commercial negotiated-rate disclosures across 115 billing codes and 9 payers, from a rate extract read on 27 August 2026, plus the 2026 Medicare Physician Fee Schedule and Clinical Laboratory Fee Schedule, Medicaid National Correct Coding Initiative edits, Medicare Local Coverage Articles, and hospital price-transparency filings from seven filers. Every source is named with the date it was read.

How far back does the preserved history go?

Collection began in July 2026, so as of September 2026 that is roughly two months of vintages, growing by one a month. It is not a multi-year archive and we do not describe it as one. What it already supports is a comparison nobody else can run: between two dated extracts of the CMS coverage record, 22 policies were revised with no changelog published.

Can I reproduce an answer months later, after the source file has been replaced?

That is the point of the vintage. Payers republish transparency files monthly and no federal rule requires the prior version to remain available, so an answer that cites a live endpoint stops being checkable the moment the endpoint moves. Answers here name the vintage they were served from, and the preserved file behind that vintage is kept and re-runnable.

Do you cover my payers and my codes?

The public world answers 115 billing codes across aetna, anthem, bcbs-mn, centene, cigna, hcsc, highmark, independence-blue-cross, unitedhealthcare. If your code family or your payer is outside that, say so in a pilot enquiry: extending the corpus for a specific code family and market is exactly what a private world is.

How is this different from a policy-tracking subscription?

Citation is the floor, not the headline; several products will hand you a cited answer. The question to ask any of them, including us, is whether it can reproduce the rate, the policy, the source version and the resulting claim together, as they stood on a date in the past. Preserved dated vintages and a refusal system are what that requires.

Does our own data have to leave our environment?

No. Nothing here needs your claims, your contracts or any PHI. The corpus is public record. If you choose to share your own published assumptions so a claim can be checked against them, that is your call and it is covered in writing.

Free is a real tier, not a trial: the current vintage, rate limited, no card. A pilot is for the case where the corpus has to include something we do not hold yet, which is most real contracting work.

Start a pilot, $20k → Pricing → Docs →

Vintage AUG 31, 2026, build 13. 25 tools that answer a question about the corpus, 43 on the wire in total.