Data report
Peptide COA report, October 2026
The index holds 648 certificates from 26 vendors. Among those whose assay was run in the last 90 days, the median gap between the labelled mass and the assayed mass is 4.80%, against 6.12% for older assays — the clearest sign in this corpus that documentation is tightening. That comparison is made on the date the laboratory ran the test, which is the only date here our own crawl schedule cannot move; the month-over-month figures this report was built to lead with do not survive scrutiny, and the next section says why.
Published 1 Oct 2026 · baseline 16 Sept 2026 · 648 certificates · every figure computed from the live index
Edited by Slava Melikov, Research Editor · reviewed 1 Oct 2026
The corpus as it stands
| Measure | Value | Reading |
|---|---|---|
| Certificate records | 648 | across 26 vendors |
| Compounds covered | 45 | |
| Records naming a checkable lab | 605 | 93% of the corpus |
| Named laboratories | 13 | |
| Median assayed purity | 99.75% | |
| Median gap to label | 5.10% | absolute, labelled against assayed |
| At or above labelled mass | 524 (81%) | |
| More than 10% under label | 21 |
By the date the laboratory tested
Every certificate carries the date its assay was run. That date is the laboratory’s, printed on the document, and it moves for nobody — unlike the date we happened to read the file. Split on it, the corpus says something a month-over-month count cannot:
| Assays | Certificates | Median gap to label | At or above label | More than 10% under |
|---|---|---|---|---|
| Run in the last 90 days | 344 | 4.80% | 85% | 8 |
| Older | 304 | 6.12% | 76% | 13 |
Recent assays sit closer to their label and more of them meet it. Two things it does not prove: a laboratory result is not a random sample of what ships, and a vendor that has started testing recently is over-represented among the recent rows. It is the strongest temporal signal this corpus supports, which is a smaller claim than it first reads as.
Why there is no month-over-month count here
This report was built to compare the corpus against itself a month ago, by filtering on the date each record entered the index. That method does not work, and it is worth saying plainly rather than quietly dropping the table.
A certificate is only written to the index when its vendor is analysed, and analysing a vendor replaces that vendor’s entire record set. So a document published a year ago re-enters dated today, leaving the “before” side and joining the “after” side in one motion. Of the 287 records this month would have been reported as additions, 100% belong to the 10 vendors re-verified since 16 Sept 2026 — Next Gen Peptides (114), PeptidePlugs (70), Amino Club (23). That is not a finding about the market. It is the shape of our own crawl schedule, and on a monthly re-verification cycle it will be true every month.
The direction of the error flatters us, which is the reason to lead with it: a backlog being read looks identical to an industry improving. The index does not retain the date a vendor first published a document, so the honest position is that certificate growth month over month is not something this dataset can measure. What it can measure is above — the laboratory’s own date — and the state of the corpus as it stands.
The severe shortfalls
21 certificates report an assayed mass more than ten percent below the mass on the label. 6 of the 287 certificates added this month falls into that band: Melanotan I lot HYDROCP7QHN3, labelled 10 mg and assayed 8.5 mg; Melanotan II lot HYDRO11WNG8L, labelled 10 mg and assayed 8.18 mg; MOTS-c lot M0022, labelled 10 mg and assayed 8.95 mg; AHK-Cu lot 155-01-DK, labelled 5 mg and assayed 4.36 mg; L-Carnitine lot 137-01-SM, labelled 1200 mg and assayed 218.95 mg; Eloralintide lot 105-01-RB, labelled 5 mg and assayed 4.42 mg. Against 287 additions that is a low rate, and it is the rate rather than the count that carries information: a corpus that grows will accumulate shortfalls whether or not the market is getting worse.
A shortfall is not fraud and this report does not call it one. Lyophilised peptide is weighed before it is filled, some is lost to the vial wall, and a certificate reporting a gap is a vendor publishing an inconvenient number rather than hiding it. What it means for a buyer is arithmetic: at ten percent under, a ten-milligram vial is a nine-milligram vial, and the price per milligram everyone quotes is understated by the same margin.
Laboratories
KMD Analytical (KMD Group LLC) (22 certificates) is new to the index this month. A laboratory appearing for the first time is a vendor changing who tests its material, which is worth more attention than a purity figure moving a hundredth of a percent.
| Laboratory | Certificates | Share of corpus |
|---|---|---|
| Freedom Diagnostics Testing | 272 | 42% |
| ILS Laboratories | 167 | 26% |
| MZ Biolabs | 31 | 5% |
| AV Analytics | 29 | 4% |
| KMD Analytical (KMD Group LLC) | 22 | 3% |
| Kovera Labs | 20 | 3% |
| Vanguard Laboratory | 18 | 3% |
| BTLabs | 17 | 3% |
| Kovera Labs and/or ILS Labs | 13 | 2% |
| Horizon Analytical | 8 | 1% |
The newest assays
| Assayed | Lot | Purity | Label → assayed | Laboratory |
|---|---|---|---|---|
| 25 Sept 2026 | AS-SEMA10 | 99.73% | 10 → 10.52 mg | Kovera Labs and/or ILS Labs |
| 24 Sept 2026 | HR-2026-1289 | 99.35% | 10 → 10.61 mg | ILS Laboratories |
| 24 Sept 2026 | AS-MS20-2 | 99.91% | 20 → 20.72 mg | ILS Labs |
| 24 Sept 2026 | HR-2026-1256 | 99.93% | 10 → 10.28 mg | ILS Laboratories |
| 23 Sept 2026 | bpc10-10 | 99.73% | 10 → 10.48 mg | ILS Laboratories |
| 23 Sept 2026 | pt10-04 | 99.93% | 10 → 10.4 mg | ILS Laboratories |
| 23 Sept 2026 | epth50-08 | 99.17% | 50 → 50.94 mg | ILS Laboratories |
| 23 Sept 2026 | ghk-15 | 99.92% | 50 → 52.35 mg | ILS Laboratories |
Questions
Did peptide purity improve between September and October 2026?
Not measurably, and purity was never the informative number. Median assayed purity across the 648 certificates here is 99.75%, and among assays run in the last 90 days it is 99.79% against 99.71% for older ones. Purity describes the share of what is in the vial that is the peptide, not how much is there. The mass figures separate vendors far better.
Does a peptide vial usually contain what the label says?
More often than not, and the share improved this month. 524 of 648 certificates (81%) report an assayed mass at or above the labelled mass, up from 271 of 361. The median absolute gap is 5.10%. But 21 certificates still report more than 10% under label, 6 more than in September.
Which laboratories test research peptides?
13 named laboratories appear across these certificates. Freedom Diagnostics Testing signed 272, ILS Laboratories signed 167, MZ Biolabs signed 31. KMD Analytical (KMD Group LLC) is new to the index this month.
Method and scope
Every figure here is computed from the live index when the page renders, so this report cannot assert something the record no longer supports. Where it compares, it compares on the date the laboratory ran the assay, which is printed on the certificate and is independent of when we read it.
It deliberately does not compare the corpus against itself a month ago. The index records when a certificate entered it, not when the vendor published it, and re-analysing a vendor rewrites all of that vendor’s records with today’s date — so a month-over-month count measures our crawl schedule. The section above sets out the arithmetic. Restoring that comparison needs the index to retain a first-seen date per document, which it does not yet.
Scope: 648 certificates from 26 vendors across 45 compounds, restricted to vendors currently published. 43 name no laboratory we can identify and are counted in the totals but excluded from the laboratory table. Every certificate has its own page, and the whole set is downloadable as CSV. A certificate describes the sample it names and nothing else.