The same molecule, prescribed for the same purpose, is advertised across this roster at figures that differ by more than fiftyfold. That is not a single market behaving strangely. It is several different things being priced under one label, and separating them is most of the work. What is actually in the vial being priced is set out in the contents article.
The spread, and what it is a spread of
Across the 459 advertised compounded semaglutide injection prices recorded here, the 10th percentile is $99 a month, the median is $175, and the 90th percentile is $280. Twenty-one rows sit under $80 and two sit above $1,100. The compounded tirzepatide injection field, 415 rows, runs $179 at the first quartile to $299 at the third, with a median of $240. Compounded oral semaglutide, 83 rows, runs from $79 to $599.
Every one of those is an advertised figure — what a seller publishes, not what a walked checkout produced. That distinction is the first reason two numbers for one molecule disagree, and it is the reason the methodology keeps the two states apart rather than promoting one into the other.
Both ends of the range are a different kind of number
Six rows in this roster advertise $25 a month, and every one of them is conditioned on commercial insurance rather than offered as a cash price. Their own notes say so: one records a maximum saving of $100 a month, another that a separate membership is still required, another that Medicare and Medicaid are not eligible. Those are the terms of a manufacturer copay program, described in the savings-card article, reaching the shelf as though it were a shelf price.
Nine further rows carry a figure that is not a medication price at all: a consultation fee, a program membership, or a start fee, with the drug billed separately and priced nowhere. At the other end, the two highest rows carry notes describing a bundle or a cash price that insurance may reduce.
Set those aside and the fiftyfold range collapses. The middle 80% of the field spans $99 to $280 — a 2.8-fold spread. That is still a large disagreement about one molecule, and it is the disagreement worth explaining. The fiftyfold version is mostly a disagreement about what is being sold.
503A and 503B are two different regulatory objects
A compounded preparation can come from a state-licensed pharmacy operating under section 503A of the Federal Food, Drug, and Cosmetic Act, or from an outsourcing facility registered under section 503B. The FDA states that drugs from outsourcing facilities must meet current good manufacturing practice requirements, that those facilities register with the agency and are inspected on a risk-based schedule, and that compounding under 503A is not subject to those manufacturing requirements and sits primarily under state oversight.[1]
Meeting CGMP is expensive, and that cost has to land somewhere in a price. So the 503A-versus-503B split is a genuine driver of the spread. It is also one a buyer usually cannot see: the pharmacy is often not named on the page, and where it is named, the registration is not an endorsement. The FDA’s own list of registered outsourcing facilities carries the caution that the information about what each facility compounds was “provided by the registered outsourcing facility at the time of registration and has not been verified by FDA”.[2]
Underneath both routes sits the same fact. Compounded drugs are not FDA-approved, which means the agency does not verify their safety, effectiveness or quality before they are marketed.[1] The questions that do separate one compounder from another are in the vetting article.
A vial and a pen are not the same unit
The approved products are sold as pens, and the manufacturers define a month accordingly. Novo Nordisk’s self-pay page states that “One month is defined as 1 box of 4 pens of Wegovy”.[3] Lilly’s defines a one-month fill as “28-days and up to 1 single-patient-use KwikPen”.[4]
Most compounded product is dispensed as a multi-dose vial drawn with syringes, and a vial is a concentration rather than a count of doses. Two consequences follow for price. A vial can stretch further or less far depending on the dose drawn from it, so the same monthly figure buys a different amount of drug from one seller to the next. And a device costs money to make, which is part of why the pen route and the vial route do not converge on one number even before regulation is considered.
Dose-indexed pricing puts its cheapest rung on the shelf
Of the 477 providers recorded here, 148 hold one price across every dose, 36 price by dose, and 293 publish nothing that settles which they do. That last number is the largest group in the file, and it means a majority of advertised figures cannot be read as a standing rate at all.
Among the semaglutide injection rows, the dose-indexed sellers advertise a lower median than the flat-price sellers — $160 against $175. The samples are 31 and 146, so the gap is not large and is not being offered as a finding about value. The structural point is the one that matters: a seller who prices by dose advertises the figure for the rung a patient occupies for a few weeks, while a flat seller advertises the figure they pay all year. A board ordered on advertised price therefore rewards the first structure by construction, which is why the flat-pricing board exists as a separate view.
A membership is a second price on a second clock
Thirty-five of the semaglutide injection rows carry a note recording a membership alongside the medication, and their median advertised medication figure is $139 — below the roster median, before the membership is added. That is the structure doing exactly what it is shaped to do: move money out of the number a comparison page reads.
The membership itself then runs on its own cadence, which may be monthly, annual or charged once at signup. Working the result into a single comparable figure is arithmetic rather than judgment, and it is set out in the annual-cost article. Where a page states a membership and declines to state the medication price, no comparable figure exists at all.
The route carries its own premium, and it is not small
Removing the needle costs money in both directions. Across the 83 compounded oral semaglutide prices recorded here the median is $199 a month, against $175 for the injection — a $24 premium at the median. For tirzepatide the gap is larger: 43 oral rows with a median of $299, against $240 for the injection.
Two different things are being paid for there, and only one of them is convenience. A monthly figure cannot distinguish a genuine formulation cost from a positioning decision, and the two routes are not interchangeable on the published evidence either — which is the subject of the route article rather than of a price table.
Price is not evidence about the preparation
Nothing in a monthly figure discloses which pharmacy compounded the product, whether the facility is 503A or 503B, what was tested, or what beyond-use date the preparation carries. A high price is not a proxy for any of it, and a low one is not proof of its absence.
The marketing around the price is not a reliable guide either. A cross-sectional study of the Colorado direct-to-consumer market identified 93 business websites advertising compounded GLP-1 products across 188 physical locations. Of those, 41 of 93 referred to FDA approval when describing compounded products and 5 of 93 described them as “generic”.[5] Both descriptions are wrong about the regulatory status of the thing being sold, and neither correlates with what it costs. How the approved and compounded routes actually differ is in the comparison article.
What the spread does and does not license
It licenses shopping. A 2.8-fold range across the middle of the field is real money, and a figure at the 90th percentile is not buying a verified advantage over one at the 10th. Applied to the published trial results, that same range moves the cost of a pound of weight loss from $45 to $127, which is worked through in the cost-per-pound article. Where a monthly quote sits against every other published figure here can be checked in the price tool, and the ordered field is the price board.
It does not license the inverse inference. Reading a low price as a warning is as unsupported as reading a high one as assurance, because the price is generated by billing structure, sourcing route and marketing position, and those are three things the buyer can ask about directly. The number on the page answers none of them.