How to evaluate a peptide vendor

Start with evidence you can check yourself

A vendor's website tells you how that vendor wants to be seen. That's what it's for.

Evaluating one means ignoring most of it and going looking for the things a third party can confirm: lab records that trace back to the lab, batch numbers you can match to a product, and prices you can line up against a competitor on the same terms.

Everything a vendor says about itself is a claim, and claims are not evidence. Our testing ledger and price market exist to make the checkable parts easy to find.

The strongest signal: a lab record you can open

A certificate of analysis tied to a specific batch, hosted by the laboratory that ran it.¹ That's the top of the evidence ladder, and very little else comes close.

A screenshot doesn't qualify, because a screenshot can be edited and usually can't be traced to anything.

Neither does one "typical result" standing in for a whole catalogue. That's one vial being asked to speak for hundreds.

Every one of the 117 records in our ledger links back to its original lab document, which is the standard worth holding vendors to. That ledger is built on openly licensed COA data, so you can audit the underlying set rather than taking our word for it either.³

For how to actually read one of these once you've found it, we've written a guide to reading a certificate of analysis. The short version: it describes one sample on one date, and nothing more.

Treat a missing test as unknown

Most peptide certificates report identity and purity. That's two questions out of several.

Sterility, endotoxin, microbial contamination, and residual solvents usually aren't on the document at all.

A test that wasn't run is not a pass. It's a blank, and a blank should stay blank in your head rather than quietly filling in as "fine."

The same applies within the tests that were run. A high purity number doesn't tell you the vial holds the mass on the label, and in our ledger there are samples above 99% purity that were more than 20% off on dose.

Six vendors can share one lab result

This one we found by accident, and it changes how you should count testing.

We checked our 112 fully documented ledger records for duplicate values. They contain only 93 distinct results.

Seven result sets appear under more than one vendor name, accounting for 26 of the 112 records. Three of those sets each appear under the same six storefronts, with identical purity and identical measured mass, for three different compounds.

There are innocent readings, and we're not going to pick one from certificates alone.

Six storefronts might all be sourcing from a single manufacturer and honestly pointing at the same underlying batch. They might be one operation running six domains. A single test might be getting reused as marketing by several resellers. The open dataset might contain duplicates.

What's certain is the practical consequence: counting vendors who "have testing" overstates how much independent testing exists. Six sellers showing the same numbers is one test, not six.

So add a step. When two vendors both show you a certificate, compare the actual values. If the purity and the measured mass match to three decimal places, you're looking at one result twice, and it tells you nothing about either vendor's own supply.

Compare prices only after the units match

Here's what the peptide market actually looks like. As of our 25 July 2026 check, we track 706 individual offers across 60 compounds and 28 vendors.

An 8x spread on MOTS-c looks like a quality signal until you look closer. These are package prices, and packages differ: different vial sizes, different quantities, different shipping origins.

So a headline comparison is close to meaningless. A sticker price only means something once the units match: cost per milligram, same quantity, same observation date.

What the spread does tell you is that this market has no price discipline, which means you can't infer quality from price in either direction. A cheap vial isn't evidence of a bad product and an expensive one isn't evidence of a good one.

Our price market normalizes and dates every observation so you can compare like with like.

Red flags

No third-party testing at all, or testing that can't be traced back to a lab record. This is the most common one by a wide margin.

One COA presented as proof for an entire product line. One sample, one date, one batch. That's the whole scope of the document.

Medical or dosing claims. A supplier of research materials that starts describing outcomes has moved into selling treatment, which is both a regulatory problem and a signal about how the business thinks.²

"Pharmaceutical grade" and similar phrases. Not a regulatory term. It means whatever the seller wants it to mean.

A price that's far outside the range for that compound, in either direction, with no explanation of what's different about the package.

The short checklist

Run these in order and stop when one fails.

1. Is there a lab record, and can you open it on the lab's own site?
2. Does it name a batch that matches the product you'd be buying?
3. Are identity, purity, and quantity reported separately?
4. Which tests aren't there?
5. Once you convert to cost per milligram, how does it sit against the range?

None of this tells you whether the compound itself is worth taking. That's a separate question, and we cover it in why evidence maturity beats online momentum and in the research directory.