Datasets
The datasets
Five datasets built from one body of data. Pick by the question you need answered. Every one lists its price, its coverage and its full column set. Three are available today and each ships a free sample you can download without asking; the other two are in preview and you can ask to join.
Five datasets
What each insurer pays for every specialty and infusion drug, and for administering it.
- CoversSpecialty and infusion drug codes, plus the administration codes
- Insurers37 insurers, 50 states and DC
- Compared toMedicare, ASP+6% and drug acquisition cost
- UpdatedEvery quarter
Questions it answers
- What does this insurer pay for zoledronic acid in Texas, and how does that compare with Medicare?
- Which of our top ten drugs are paid below the federal benchmark, and in which states?
$16,000a year · slices from $6,000
The same rates across all 350,000+ billing codes, not just drugs.
- CoversAll 350,000+ billing codes
- Insurers37 insurers, 50 states and DC
- Compared toMedicare and, for drugs, ASP+6%
- UpdatedEvery quarter
Questions it answers
- What does each insurer pay for a knee replacement in our metro area?
- Where does our current fee schedule sit against the market, line by line?
$60,000a year for everything · slices from $8,000
Every hurdle an insurer puts between a prescription and a payment — prior authorization, step therapy, site limits, renewals.
- Size2,423 drug-and-insurer pairs — 104 drugs across 30 insurers
- Built from2,634 insurer policy documents, read and quoted
- CoversPrior authorization, step therapy, site of care, renewals
- Depth81% of pairs require prior authorization
Questions it answers
- Which insurers require two other drugs to fail before ours, and which two?
- What did this insurer’s policy actually say in March, when the claim was denied?
$150,000a year for everything · slices from $6,000
Rates and coverage rules in the same row. No other published dataset has both.
- CombinesNegotiated rates and coverage rules in one row
- Gives youA friction score and a flag for underpaid, hard-to-get drugs
- TraceableEvery row links back to the exact quoted rule
- AvailableNowhere else — no other published dataset has both halves
Questions it answers
- Rank every insurer by what they pay for our drug and how hard they make it to get.
- Where is our drug both underpaid and sitting behind a competitor in step therapy?
$210,000a year for everything · slices from $20,000
What one named practice is contracted for, with which insurers, at what rates — down to the plan.
- Identity18.6M tax-ID-to-provider links across 890,761 resolved practices
- Names1.47M alternate and legal names, so one practice is one row
- Insurers acceptedWhich insurers each practice holds a contract with, per provider
- Grounded byMedicare volume, so a dead contract line is visible
Questions it answers
- We are buying this infusion group. What are its contracted rates, and how do they compare with the market?
- Which practices in this metro have the strongest commercial contracts for our drug?
$80,000a year for everything · slices from $10,000
Why there is no “by insurer” or “by state” dataset
Those are filters, not products. By drug, by insurer, by state and by setting are all the same table with a different filter on it, and selling them separately would mean buying the same rows twice. Narrowing the scope lowers the price. It does not change the dataset.
If you need one insurer, in one state, for one drug family, say so and we will price that. If you need all of it, that is priced too.
Compare them side by side
| Specialty Drug Rates | Complete Rate File | Payer Coverage Requirements | Rates + Requirements | Contracted Rates by Provider |
|---|
| Answers | What each insurer pays for every specialty and infusion drug, and for administering it. | The same rates across all 350,000+ billing codes, not just drugs. | Every hurdle an insurer puts between a prescription and a payment — prior authorization, step therapy, site limits, renewals. | Rates and coverage rules in the same row. No other published dataset has both. | What one named practice is contracted for, with which insurers, at what rates — down to the plan. |
|---|
| Price | $16,000
a year · slices from $6,000 | $60,000
a year for everything · slices from $8,000 | $150,000
a year for everything · slices from $6,000 | $210,000
a year for everything · slices from $20,000 | $80,000
a year for everything · slices from $10,000 |
|---|
| Status | Available now | Available now | Available now | In preview | In preview |
|---|
| Free sample | Yes | Yes | Yes | Not yet | Not yet |
|---|
| Table | rate_intel | rate_market | coverage_cell coverage_rule | rate_intel coverage_cell coverage_rule | rate_intel practice practice_member |
|---|
| Updated | Quarterly | Quarterly | Quarterly | Quarterly | Quarterly |
|---|
Why the prices are what they are
Every number below is anchored to something you can check — a competitor’s published list price, or an awarded public contract. None of it is a guess dressed up as a rate card.
How a slice is priced
A slice that covers a share of the grid costs the square root of that share, times the full price.
Half the data is not half the work. Building, checking and delivering a narrow slice costs nearly what the whole thing costs, and a slice is worth more to the person who needs exactly it than its size suggests. A straight percentage would price small slices below what they cost us to ship and large slices above what they are worth to you.
Narrow the codes
One molecule, one drug family, one therapeutic area, or the specialty and infusion set instead of all 353,270 codes.
Narrow the market
One insurer, one state, or a defined region instead of 37 insurers across 50 states and DC.
Narrow both
Multiply the two shares together, then take the square root of the result.
One therapeutic area, in one region
- about 14% of the drug set
- about 20% of the market
- 14% × 20% = 2.8% of the grid. The square root of 2.8% is 17%.
17% of $75,000 is $12,600.
No slice prices below $6,000. Below that the delivery, the manifest and the support cost more than the licence, and we would rather say so than quietly ship you something worse.
Slices stack and never double-charge. If you buy one drug family and add a second later, you pay the difference between the two rungs, not the second rung again.
One thing the arithmetic does not capture: the specialty and infusion set is 5,550 codes out of 353,270 — 1.6% of the code universe and 6.8% of the rows — but it is priced at roughly a quarter of the full file. That is deliberate. Those codes carry the decisions that are worth money. We price a slice by how much of the question it answers, not by how many rows it has.
Rates undercut the market on purpose
Other vendors sell negotiated rates built from the same public filings, so this half of the catalogue competes on price. The only published price in the category is a competitor listing one insurer’s rate file at $25,000 a year. The same thing here is $18,000. The full national file is $60,000, at the bottom of the $42,000 to $71,000 band that awarded federal contracts put comparable healthcare databases in.
Coverage Requirements is priced as the premium it is
No rate vendor sells this at all, and the vendors that do sell payer policy publish no prices and sell into enterprise pharma budgets. It is also by far the most expensive thing here to produce — every rule is read out of the insurer’s own document and quoted, never scraped or inferred. $150,000 for the full corpus, and it is the one dataset we do not discount to win a deal.
The join at $210,000
Large national claims databases run $190,000 to $233,000 a year on awarded federal contracts. Rates joined to the rules that gate them is the only dataset of its kind, and it is priced inside that band rather than above it.
Every slice is priced by a published rule
A slice covering a share of the grid costs the square root of that share, times the full price. That rule is written out with a worked example, so you can price your own subset before you contact us. Where the market says something different from the rule — the single-insurer rate file is the one real case — the market wins and we say so on the line.
Every slice is a real subset
A slice is the same columns, the same quality gates and the same quarterly refresh over less of the grid. Nothing is held back to make the full set look better, and slices stack — you are never asked to re-buy what you already hold.
What every dataset includes
Parquet and CSV
Files are split by quarter and state, so you can load only the part you need.
Quarterly updates
Delivered on the insurer publishing cycle. Each delivery includes a manifest with row counts, coverage and the schema version.
A schema that holds still
Columns are versioned and every change is written down in the changelog, so what you build on top keeps working.
Everyone at your company
Annual license, unlimited internal users, no charge per query. Terms for putting it inside a product you sell are on the trust page.
What a delivery looks like
Files are split by quarter and state. Every delivery carries a manifest with the row counts, the coverage and the schema version, so you can check what arrived before you load it.
manifest.json — ships with every delivery
{
"dataset": "infusion-rate-benchmark",
"vintage_quarter": "2026-07-01",
"schema_version": "1.2.0",
"generated_at": "2026-07-14T09:12:04Z",
"files": [
{ "path": "state=TX/part-0000.parquet", "rows": 1284119, "bytes": 41205336 }
],
"coverage": {
"insurers": 18,
"states": 51,
"billing_codes": 1146,
"rows_total": 36914552
},
"quality": {
"confidence_score_median": 84,
"rows_below_70": 0.061,
"outlier_flagged": 0.004
}
}
Not sure which one you need
Tell us the question. We will point you at the smallest dataset that answers it, or just run it for you.