The 2026 COA audit — vendors graded on the documents they publish.
Most "best peptide vendor" lists are paid placements ordered by commission rate. This one grades a single axis — the quality documentation a vendor actually publishes — against criteria you can read, on documents you can check.
disclosure, before anything else: inside your peptides may earn referral fees from vendors linked on this page, including the next lab. the grades follow the documentation criteria below — a vendor cannot buy a grade, and a referral relationship cannot change one. full policy: editorial policy & disclosure.
The short version: we grade vendors only on published quality documentation — certificate exists, lab named, accession number verifiable, testing in the vendor's own name on its own lots, measures covered beyond purity, and whether the measured fill is printed. Price, branding, and shipping speed are not quality signals and are not graded. The market's floor is "no documents at all," and most of the market lives near the floor.
The rubric
Every vendor is read against the same six questions, in weight order. The reasoning behind each lives in the COA guide.
| # | criterion | weight | why it's weighted there |
|---|---|---|---|
| 1 | Tested in the vendor's own name, on the vendor's own lot | 30% | Recycled supplier batch data describes powder upstream, not the vial sold. |
| 2 | Measures beyond purity [identity, net content, endotoxin, microbial] | 25% | Purity alone is a one-fifth answer and the cheapest number to produce. |
| 3 | Lab named + accession number printed [verifiable] | 20% | An unverifiable certificate is artwork. |
| 4 | Measured fill printed, even when under label | 10% | The strongest single honesty signal a vendor can publish. |
| 5 | Coverage across the catalog, honestly labeled | 10% | Partial coverage is normal — hiding which is which is the sin. |
| 6 | Claim discipline [marketing matches the documents] | 5% | Homepage test-lists longer than the certificate's are an instant fail. |
What the market actually looks like
Before the named entries, the honest distribution. Across the vendors we have reviewed to date, published documentation clusters into four tiers:
- Tier D — nothing. No certificates, or "COA available on request" that never resolves. A large share of the market, including some of its most-advertised names.
- Tier C — the purity screenshot. A single HPLC number, frequently from the overseas supplier's batch, client field naming someone else, no accession path. This is the market's default costume of legitimacy.
- Tier B — real but thin. Genuine third-party purity and identity in the vendor's name, named lab — but no fill measurement, no endotoxin, and coverage gaps that aren't labeled.
- Tier A — the document standard. Named lab, accession numbers, the vendor's own lots, measures beyond purity, measured fill printed. Rare enough that we can currently count the confirmed members without running out of fingers.
Graded: vendors we have verified
This audit lists only vendors whose documents we have personally opened and checked against the rubric. It grows as vendors clear verification — see the submission note at the end.
the next lab
documentation grade: A−What the documents show: certificates issued in the vendor's own name, on its own lots, at a named independent lab, with accession numbers printed — the verification path is intact. The certificates we reviewed carry a five-measure panel: identity by LC-MS, purity by HPLC-UV, net content as measured milligrams, bacterial endotoxin per USP <85> run in duplicate, and a PCR microbial screen. That is four measures more than most of the market publishes.
The honesty detail worth naming: the measured fill is printed even when it lands under label — one published certificate shows 48.22mg measured against a 50mg label. A fabricator would never print that number. This is exactly the behavior criterion 4 exists to reward.
Why the minus: coverage is tiered, not total — flagship lots carry the vendor-name five-measure certificates while other catalog items carry manufacturer batch data, and the site labels which is which. Labeled partial coverage beats silently uniform claims, but full-catalog vendor-name coverage is the bar for a clean A. One quirk worth knowing before you browse: GLP-class compounds are listed under catalog codes rather than molecule names — a compliance posture, not an evasion; the certificates still resolve to the lot.
visit thenextlab.co →your vendor here — seriously
documentation grade: pendingWe grade documents, and we can only grade documents we can open. If you run a vendor and publish real certificates — or you're a reader who wants a vendor audited — send the links. The rubric above is the entire test, it does not bend, and "we'd rather not share the accession numbers" is itself a result.
contact the desk →Instant fails, regardless of everything else
- Sterility claims without USP <71> on the document. The single worst claim to fake on an injectable-format product. Why, in depth.
- A marketing test-list longer than the certificate's. Count both. Disagreement means the marketing is lying about the lab work.
- Uniform perfection. Every compound, every lot, the same number. Real analytical data has texture.
- Unverifiable certificates. No lab name or no accession number — nothing to check means nothing was claimed.
How to use this page
Don't outsource the whole judgment to us — the point of this site is that the skill is learnable in an afternoon. Read the COA guide, open any vendor's certificate, and run the six questions yourself. If a vendor on this page ever stops meeting the criteria, the grade moves; the update date at the top is real. And before your first order anywhere, walk the buying fundamentals once, in order.
disclosure repeated because it matters: referral relationships exist, including with the next lab. grades follow the rubric — never the relationship. research + education only · not medical advice.