Provider Price Index

What this page assumes about you

That you are paying cash, that you will hold a maintenance dose rather than a starter dose, and that a difference of a few hundred dollars across a year is worth an hour of reading. If any of those is wrong, the ordering here changes.

Insurance is the biggest one. A covered prescription under a documented indication beats every cash route on this site, and establishing whether you qualify comes before comparing 20 cash prices spanning $145 to $324.

The bias we can see in our own data

We track what programmes publish, so programmes that publish well look better here than programmes that treat pricing as a sales conversation. That is a real bias and we would rather name it than pretend the dataset is neutral.

It cuts a defensible way — a programme unwilling to state a price before an intake has made a choice you should notice — but it is a bias, and 12 tracked programmes appear here with an explanation instead of a number because of it.

What we deliberately do not measure

Shipping reliability, response times, and whether the clinical oversight is any good. None is observable from outside without enrolling, and we did not enrol.

That absence is why there is no rating out of ten anywhere here. A single score would compress price, disclosure, service and clinical depth into one figure and hide the weighting — which is precisely the trick that makes comparison sites feel authoritative while telling you less than a table would.

The usable proxy is what a programme publishes before it has your money, and that is what every disclosure column here records.

The switching cost nobody prices

Moving programmes for a modest saving carries two costs a table cannot show: a supply gap while a new intake is reviewed, and a new prescriber restarting titration rather than continuing your dose.

The second is expensive. Sixteen weeks back through the ladder erases most of what a year's saving would have bought. Ask for dose continuation in writing before cancelling anything, and do not cancel until the new programme has shipped.

Method

Computed over one file covering 32 programmes on a single price basis: all-in monthly cost, medication plus any recurring fee, at a named dose tier. No estimates, no imputed values, and no programme excluded for reasons other than the criterion stated.

Where a finding depends on a figure we could not verify at source, it is flagged rather than rounded away.

Findings

All-in cost at a 2.4 mg maintenance dose spans $145 to $324 — a 2.2-times spread for the identical molecule from the same category of licensed pharmacy. 19 of 20 programmes hold one price at every dose; the rest reprice as you climb. 4 charge a mandatory recurring fee on top of medication.

Fewer than a fifth identify the dispensing pharmacy before purchase, which is the single disclosure that would let a patient verify anything independently.

Limitations

Everything here comes from published material we could access. We did not enrol, did not mystery-shop, and did not contact programmes for comment before publication. Service quality, shipping reliability and clinical depth are absent because they are not measurable from outside.

A programme that discloses something we could not find should tell us, and the record will be corrected with the date.

Reuse

Cite freely with attribution and the capture date. Journalists and analysts should take the underlying file rather than transcribing from a table, because the file carries the evidence status that a table necessarily compresses.

The short version of provider price index

Price the dose you expect to hold, not the one in the advertisement. Add every recurring fee. Verify the pharmacy. Everything else on this page is detail underneath those three.

Done properly the answer is usually one of a handful of programmes: the cheapest verified route runs $145 a month all-in, about $1,740 across a first year including titration.

The next step that actually moves things

Send two or three programmes the same five questions: total at 2.4 mg including every fee, which pharmacy, is the prescriber licensed in my state, what notice to cancel and what is refundable, and which form of the active ingredient is used.

Whoever answers all five in writing has told you more than any comparison table can, this one included. Whoever does not has also told you something.

One thing worth doing annually

Re-price the market at your actual maintenance dose. No programme notifies existing patients when a competitor drops below it, and semaglutide pricing moved repeatedly through 2025 and 2026 — manufacturer cuts, an oral option, a higher-dose product, and several platforms leaving compounded entirely.

An hour once a year is a poor use of time for most purchases and an excellent one for a medication you may take for years.

What is in the file

32 programme records. Each carries name, type, pricing model, medication cost, any recurring fee, all-in cost at five dose tiers, a first-year total, cost per milligram, prepaid rate and term, published state count, care-model attributes, evidence status, source text, a resolvable source URL, a Wayback lookup and the capture date.

20 of them carry a price. The remainder carry an explanation of why not, which is a field rather than a silent null.

How to work with it

The JSON is the canonical form and carries claim-level citations; the CSV is flattened for spreadsheets and drops the nested care-model and claims objects. Both are regenerated on every build, so a figure in the file and a figure on a page cannot drift apart.

Sort by first_year to reproduce our cheapest ranking. Filter evidence_status to source-verified for the subset we would defend. Group by pricing_model to see the flat-versus-scaled split.

What it deliberately omits

Ratings, scores out of ten, review counts and sentiment. None of those is measurable from public material in a way we would defend, and publishing them would make the file look more authoritative than it is.

It also omits estimated prices. A programme that does not publish appears with nulls and a reason, because a null is checkable and an estimate is not.

Licence and citation

Reuse permitted with attribution to this site, carrying the capture date shown on the figure. A price without its date is not a fact, and republishing one without it is how the errors we spend our time correcting get created in the first place.